Re: [PyMOL] interior protein-surfaces

[DeLano S. <de...@de...>]

One pure PyMOL-based approach is to use the surface_carve settings as
follows:

load $TUT/1hpv.pdb, tmp
extract lig, organic
extract prot, polymer
delete tmp

set surface_carve_cutoff, 4.5
set surface_carve_selection, lig
set surface_carve_normal_cutoff, -0.1

show surface, prot within 8 of lig
set two_sided_lighting
set transparency, 0.5
show sticks, lig
orient lig

This approach culls away any triangle more than 4.5 angstroms away from all
ligand atoms and any triangle not facing at least one ligand atom.  It tends
to produce a fairly clean pocket representation.

Also worth trying (after the above):

set surface_color, white
set surface_type, 2
unset ray_shadows

Cheers,
Warren

--
DeLano Scientific LLC
Subscriber Support Services
mailto:su...@de...
 
________________________________

From: pym...@li...
[mailto:pym...@li...] On Behalf Of Thomas S.
Leyh, Ph. D.
Sent: Tuesday, May 27, 2008 1:53 PM
To: pym...@li...
Subject: [PyMOL] interior protein-surfaces


     From time to time I find it valuable to view the suface of a
ligand-binding pocket looking out from the surface's interior.   This is
accomplished in a variety of ways - principally, zooming and clipping.  It
would be wonderful to be able to represent just the "skin" of the active
site wrapping around the ligands, like a sac - this seems quite difficult to
achieve.  A related problem is that ray-ing a surface that has been clipped
yields a collection of odd surface regions that represent poorly the
non-ray-ed version - it is as if the resolution of the ray it too gross to
capture the clipping edges. 
 
   Advice appreciated.
 
    Thanks,
 
    Tom Leyh