Date: Wed, 18 Dec 2002 11:56:20 -0800
From: Richard Henchman
Subject: Re: SUMMARY: solvating a protein with large cavity

Concerning other ways to solvate a protein, I've found that using GRID
is quite a good way. It essentially docks water molecules into the
protein. You need to a license, free for academics.

http://www.moldiscovery.com/manual/index.html

In summary, you run
1. GRIN to setup the system (via GREAT)
2. GRID to create the grid for water probe (via GREAT)
3. MINIM to find the minima in the grid
4. FILMAP to add waters to the grid

The only really important parameter in the method is the minimum
energy depth to define a minima (set in step 3). The version GRID 20
has a GUI called "Greater" but I've found the simpler GREAT to be
sufficient. It takes only about 5 minutes when you know what
you're doing (step 2 is the slowest and scales with protein size).

GRID only gives oxygen positions (at least, the way I was doing
it). If you really want nicely placed waters, then I use WHATIF to
place hydrogens on these waters to give the water the most favorable
orientation for hydrogen bonding. This is probably not critical
since waters can usually reorient on a fairly fast time scale.

Afterwards I run solvatebox in leap in the normal way to complete
solvation.

Richard

Tom Bishop's message on Wed, Dec 18, 2002 at 09:03:27AM -0600 was:
> For anyone that was following this thread.
> Here's what we came up with.
>
> In general the "Leap" solvate commands work well for bulk solvent, but
> do not expect these commands to place ANY waters specifically. The
> equilibration phase will
> locate waters specifically if the location is accessible, but cavities
> should be addressed
> on an individual basis, even for very large cavities.
>
> Good luck,
> TOm Bishop
>
>
> On Mon, Dec 16, 2002, Thomas wrote:
> >
> > Here's what I finally worked out to get the waters put in, where
> > {27.7 25.9 35.4} is the center of the cavity
> >
> > solvatecap REC WATBOX216 { 27.7 25.9 35.4 } 12 .75
> >
> > solvatebox REC WATBOX216 6
> >
>
> Sounds good. I have never used the solvateCap or solvateShell commands
> myself, so I can't help out very much.
> >
> >
> > In general if there are small cavities in a protein would it make sense
> > to solvate it in two stages. Stage one try to gets waters in the
> > cavities with a reducued closeness
> > parameters and then stage 2 add bulk waters.
> >
>
> Stage 1 might require some different program. Take a look at:
>
> http://www.bch.msu.edu/labs/kuhn/web/software.html
>
> A google search on "protein solvation sites" turns up a number of
> promising
> hits (including the one above). You might also ask on CCL how others
> get
> starting positions for waters in cavities.
>
> ..good luck...dac
>
> --