Re: Kb, Ktheta values for TIP3P water model

From: Maxim Belkin (
Date: Mon Aug 11 2014 - 18:05:41 CDT


Could you please copy/paste entries related to water atoms from the topology (and parameter) files you use to generate (and simulate) your system?

Since you report "unreasonable displacements" for all atoms, I can assume that you put "0" as a spring constant for the water's "OH - HT" bonds. Do not do that. Use some non-zero value for this bond.

Then, NAMD does not require the bond between hydrogens for water to be rigid - you can either use 0 for "HT - HT" bond, or remove this entry from the topology and parameter files completely. So, if you do use 'rigidBonds all' in your NAMD configuration file (AND all your inputs are correct), then your water molecules must be rigid.

Take a look:
There, Jim explains some points regarding CHARMM TIP3 water in NAMD.

One way to check bonds/angles in a molecule is straightforward: load your trajectory in VMD, label any bond or angle in a water molecule (VMD menu: Mouse -> Label -> Bonds/Angles), and then go to Graphics -> (Bonds/Angles) -> choose a bond/angle that you have just labeled and then click 'Graph...' in the 'Graph' tab. If bond length or angle of a water molecule is not constant, then there is a problem with either of your inputs.

Good luck,

On Jul 29, 2014, at 4:38 PM, Viswanath Pasumarthi <> wrote:

> Hi,
> I have been trying to study interface between bulk phases of chloroform
> and water. I am following the TIP3P model for water proposed in Jorgensen
> et al. 2003 ( In the force field
> parameters, what should be the Kb, Ktheta values be provided since TIP3P
> considers water to be a rigid monomer. If I take their values as 0, the
> system is undergoing chaos with each individual atom undergoing
> unreasonable displacements. If I take their values as high as 1000, the
> system is not minimizing.
> Thank you,
> Viswanath.

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:43 CST