Re: H-H bond in TIP3 water

From: Jim Phillips (
Date: Fri Aug 05 2005 - 11:11:15 CDT

In general, always use "auto none" to build water. Here's why...

This is a standard CHARMM TIP3 residue with three bonds and one angle:

RESI TIP3 0.000 ! tip3p water model, generate using noangle
ATOM OH2 OT -0.834
ATOM H1 HT 0.417
ATOM H2 HT 0.417
BOND OH2 H1 OH2 H2 H1 H2 ! the last bond is needed for shake
ANGLE H1 OH2 H2 ! required

If you build this without auto none you get the explicit H1-O-H2, angle,
plus three generated angles: H1-O-H2, O-H1-H2, and H1-H2-O. Since there
aren't parameters for H-H-O, NAMD dies.

If you remove the H-H bond (since NAMD doesn't use it and it looks funny
in VMD) and build without auto none you get the explicit H1-O-H2 angle,
but psfgen generates a second H1-O-H2 angle. If you use rigid water this
doesn't matter, but flexible water will have a stiller angle.

Solvate removes the H-H bond, but uses "auto none", so you get only the
explicit angle. This is fine and correct.

If you really hate "auto none", then remove both the H-H bond and the
explicit angle. You'll just get the generated angle, which is fine.

You might think that you have rigid water so you don't need the angle
either, but NAMD needs it for minimization and to know what angle to
constrain the rigid waters to.


On Mon, 1 Aug 2005, Jorge Pikunic wrote:

> Hello,
> Is "auto none" required to build water segments? or not anymore?
> According to the NAMD user's guide, it is required
> (
> "The residue type for water depends on the model, so here we alias HOH
> to TIP3. Because CHARMM uses an additional H-H bond we must disable
> generation of angles and dihedrals for segments containing water. Then
> read the pdb file.
> # (7) Build water segment
> pdbalias residue HOH TIP3 ; # formerly "alias residue ..."
> segment SOLV {
> auto none
> pdb output/6PTI_water.pdb
> } "
> But according to these posts, it is not required:
> "If you're going to be running your simulation in NAMD, it's safe to
> edit the topology file and remove the H1-H2 bond. NAMD will ignore it
> anyway, since it doesn't need the bond to do rigid waters, and your
> VMD images will look better."
> Which of the two versions is true? It seems that the user's guide is not
> updated. Or am I misinterpreting something?
> Please help :) Thank you!
> Jorge

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:39:48 CST