[gmx-users] rigid body again

Anton Feenstra feenstra at few.vu.nl
Thu Jun 16 10:25:02 CEST 2005

David van der Spoel wrote:

> On Thu, 2005-06-09 at 10:35 -0400, Lianqing Zheng wrote:
>>Thanks, David! What is long range constraint? For an i-j-k-l dihedral,
>>try to constrain i---l distance?
> no, you  just define them yourself, from one end of the molecule to the
> other.

If you want to construct your molecule as a *really* rigid body, you may 
want to consider the following. Construct a small set of point-masses 
(i.e., atoms without charge or LJ) that reproduce your molecule's center 
of mass and principal axes. You will need at least three of these 
masses, but four could be more convenient in the following steps. These 
three or four masses, you tie together with (precisely defined) 
constraints, so they have no internal degrees of freedom with respect to 
each other. Then, you can construct all the atoms of your molecule as 
'dummy atoms' where the positions are generated as linear combinations 
of the set of masses (and, optionally, an outer product). Forces will be 
re-distributed back from the (mass-less) dummy atoms to the masses.

This type of constructions have been described for hydrogens in 
biomolecules (e.g., -NH3 groups in proteins) in J. Comput. Chem. 20 (8), 
p786, 1999, and for aromatic aminoacid sidechains in J. Biomol. NMR 23, 
181-194, 2002. (And in my thesis: www.few.vu.nl/~feenstra/thesis.pdf)

For the hydrogens, these constructions are stable and work well, 
allowing timesteps of 4 to 6 fs. For the aromatics, it may be that there 
are either some inherent (algorithmic?) problems, or possible problems 
with the constructions as they are generated automatically by pdb2gmx 
(option -dummy aromatics).



* NOTE: New Phone & Fax numbers (below) *

  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 59 87608 - Fax: +31 20 59 87610           |
|             | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/         |
|             | "If You See Me Getting High, Knock Me Down" (RHCP)    |

More information about the gromacs.org_gmx-users mailing list