From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Sat May 23 2009 - 14:19:37 CDT

Hi Zizi,
consider a 10x10x10 cube. What is the surface area of only the slices of
the cube along the z axis between even and odd integers (ie, 0-1 U 2-3 U
4-5...), in relation to that of the whole cube?

The proper way to calculate the portion of the surface of a selection
which is contributed by a smaller selection (eg, the sasa of only
hydrophobic parts of a molecule) is

measure sasa srad $fullsel -restrict $hpsel

where $fullsel is a selection consisting of the entire molecule and
$hpsel is only the hydrophobic atoms. This will calculate the solvent
accessible surface for the full molecule and then only return the area
of the parts near hydrophobic atoms. If you just do
measure sasa srad $hpsel
then you're actually calculating the sasa of a bunch of isolated
clusters of atoms in space, which will be higher than the area
contributed by them in the complete molecule.

Best,
Peter

zizi moradi wrote:
> Dear all,
> I calculated the solvent accessibility for non polar and polar part of a
> nucleic acid.The SASA for hydrophibic part alone is larger from the
> SASA of the whole structure(considering polar and non polar
> together).Why is like this?
>
> thanks,
>
> Zizi
>
>