From: Josh Vermaas (vermaas2_at_illinois.edu)
Date: Tue Aug 27 2013 - 10:34:32 CDT

Hi Mohammad,

You just need to be careful with your selections. You only loaded the
pdb into your VMD session, and the pdb format doesn't specify a segname
(psf files do). Thus, if you checked the number of atoms selected:

set TMCb [atomselect top "segname TMC and chain B"]
$TMCb num

You'd get 0. The accessible surface area around zero atoms is zero, and
that's why you get the answer you don't expect. However, using:
set TMCb [atomselect top "resname TMC and chain B"]

would probably give you the answer you want. Visually inspecting
atomselections is often a very useful sanity check in these situations
to make sure you are selecting what you think you are.

-Josh Vermaas

On 08/27/2013 08:35 AM, m r wrote:
>
> Dear all,
>
> I'm calculating the buried surface area for the complex 1E2K.pdb.
> It's composed of chains A and B, with a copy of the ligand "TMC"
> bound to each of the chains. However, TMC bound to chain A is
> completely covered by the protein, while TMC bound to chain B is
> partially accessible by solvent.
>
> To check this in VMD, I used the following commands:
>
> mol new 1E2K.pdb
> set complex [atomselect top "all"]
> set n 500
>
> set prA [atomselect top "protein and chain A"]
> expr {([measure sasa 1.4 $complex -restrict $prA -samples $n])}
> 7869.8974609375
>
> set prB [atomselect top "protein and chain B"]
> expr {([measure sasa 1.4 $complex -restrict $prB -samples $n])}
> 8787.044921875
>
> set TMCa [atomselect top "segname TMC and chain A"]
> set complexA [atomselect top "(protein and chain A) or (segname TMC
> and chain A)"]
> expr {([measure sasa 1.4 $complex -restrict $TMCa -samples $n])}
> 0.0
>
> set TMCb [atomselect top "segname TMC and chain B"]
> set complexB [atomselect top "(protein and chain B) or (segname TMC
> and chain B)"]
> expr {([measure sasa 1.4 $complex -restrict $TMCb -samples $n])}
> 0.0
>
>
> The last result is obviously wrong, as the ligand "TMC" is partially
> bound and
> accessible by solvent. Due to this result (i.e. zero ASA for the ligand),
> I didn't continue to calculate the buried surface area from the
> following equation:
> [SASA(prB) + SASA(TMCb)] - SASA(complexB)
>
> I appreciate if anyone could point out why the last result (for TMCb)
> is wrong
> and how to deal with it.
>
> regards,
> Mohammad
>
>