Re: Why there is a bond between the two Hydrogens of the TIP3P water in Amber force field?

From: Peter Freddolino (
Date: Wed Oct 29 2008 - 11:07:13 CDT

Purushottam Dixit wrote:
> Hi,
> If you are using SHAKE to maintain rigid water conformations (this
> would read as 'rigidbonds water' or 'rigidbonds all'), there's a
> pseudo bond of high spring constant between the hydrogens. Other than
> maintaining the rigidity it has no function.
I'm sorry, but this is slightly incorrect. If the water is kept rigid by
rigidbonds (using SETTLE, a special constraint solver for TIP3P water)
then the bond distances and H-O-H angle are truly constrained (ie, not
allowed to vary at all from their equilibrium positions, and having no
associated potential); there is no harmonic term here as a spring
constant would indicate. The function of the H-H bond, which shows up in
water prepared using charmm or amber, is to allow those programs to
obtain proper geometries for the rigid water constraints; namd instead
calculates the geometry from the O-H equilibrium bond lengths and H-O-H
equilibrium angle. Thus, the bond doesn't generally appear in the
topology files we distribute, and will be ignored y namd.
> Purushottam
> On Wed, Oct 29, 2008 at 11:20 AM, Ming <
> <>> wrote:
> Dear all,
> I am trying to simulate my system using namd with the amber-style
> force field.
> Everything was fine! However I found there is a bond between the
> two Hydrogens of
> the TIP3P water in Amber force field. For what reason they put a
> bond between the
> two Hydrogen? Any comments? Thanks in advance!
> Best!
> Ming

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:48:32 CST