[gmx-users] defining a group within a cutoff

[Anton Feenstra]

Yunfeng Hu wrote:

> Hi, I wonder if anyone can give a clear answer of how to use trjorder 
> and/or other commands to define an index.ndx file which can be used for 
> h_bond analysis purposes. Thanks.

trjorder sorts your water (or other solvent) molecules based on distance
to a reference group. So, you can get water close to your protein first
in the trajectory, by selecting the protein as reference. Then you'll
have to see (using e.g. rasmol or vmd) how many water molecules you will
need for a certain water layer thickness around your protein. You can
select the corresponding range of residue numbers in make_ndx (use, e.g.,
'r 351-2457') or use a text editor (or script) to write out the corresponding
atom numbers.

What you'll loose in this way is the correlation between water molecules
from one frame to the next - they have been re-sorted for each frame!

Otherwise, g_hbond should be able to handle the whole bulk of water, but
may require excessive amounts of memory to build the necessary neighbor-
lists.


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610           |
|             | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/         |
|             | "If You See Me Getting High, Knock Me Down" (RHCP)    |
|_____________|_______________________________________________________|