VMD-L Mailing List

From: Suman Chakrabarty (suman_at_sscu.iisc.ernet.in)
Date: Tue May 12 2009 - 16:33:11 CDT

Axel Kohlmeyer wrote:
> suman,
>

Hi Axel,

>> to reframe my question. I have a hydrophobic molecule in water and
>> would like to show the regions around the molecule where there is no
>> water. How do I achieve this? Using volmap and volumeslice
>>
>
> what want to do is an ill defined problem. it is always difficult
> to show the absence of something. also, i would doubt whether there
> will be a significant void unless you are running an NVT ensemble
> with negative pressure.
>

Actually the molecule is a model polymer, which upon self-organization
creates ring-like structures with no water inside (those are too small
to sustain the H-bonded network). I am interested in the molecular
mechanism of such self-organization and creation of cavities in water.
Thus it would be important to see how water is moving out and/or the
spatial distribution of water around those local structures. While
VolumeSlice does just that, it shows a particular plane. But it would be
nice if we could generate the 3D representation of the patches with no
water inside (or below certain number density) throughout the polymer
chain at once.

> that all being said, the straightforward way to look into this would
> be to compute (and show) the density distribution of your hydrophobic
> molecule and the water around it with the molecule first being aligned
> to a reference orientation and then computing the density map of the
> water molecules. depending on the size and shape of your molecule,
> simply computing radial distribution functions (g(r)) relative to
> the water oxygen atoms could be sufficient, too.
>

While g(r) gives a useful average picture indeed (and we have calculated
it), since the self-organization differs along the length of the chain,
it would be useful if we could see the molecule as a whole along with
the cavity formation in water. I would like to visualize the
heterogeneity inherent to the system. Also, g(r) can be quantitatively
misleading due to the spherical asymmetry near a surface caused by
excluded volume due to the polymer.

>> representation I can show this as slices, but I could not show all the
>> regions together for the whole molecule. I have a feeling that the
>> isosurface representation can probably do this, but I could not figure
>> out how to show the regions with density range "0 to <some cut-off
>> value)" around the solute.
>>
>
> actually, an isosurface shows the the volume enclosed from zero up to
> the specified isovalue. have you considered that if you don't see what
> you expect that the reason may be that you have not computed what you
> think you are computing? this is more often than not the reason why
> people are confused about how to use a representation. you always have
> to keep in mind that VMD is programmed to be incredibly ignorant about
> what a person's intents are and just does what it gets told to do. ;)
>

I hope that I have been able to explain my need successfully above. The
problem is that perhaps I need to put the surface exactly at the
boundary, where water starts to appear. The Surf representation created
by using some probe molecule would not work since they do not consider
the sustainability of the H-bonded network inside a cavity.

Regards,
Suman. 

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.