Re: [AMBER] Making sense of tip3p.frcmod

From: David A Case <>
Date: Thu, 18 Feb 2021 21:24:39 -0500

On Fri, Feb 19, 2021, Paul, Austin wrote:
>OW-HW 553.0 0.9572 TIP3P water
>HW-HW 553.0 1.5136 TIP3P water
>Yet, TIP3P is a rigid model, so are these values even relevant?

You are correct: the force constants are not relevant for a rigid water.
If you want to use tip3p, you should always ask for SHAKE (ntc=2), which
actually becomes SETTLE for water.

Sometimes, one wants to minimize a conformation with TIP3P waters using
an algorithm that is not compatible with SHAKE/SETTLE. The the force
constants above keep the molecule "pretty" rigid. It's not longer really
TIP3P, but is often close enough for such purposes.

>What would happen if someone ran a simulation without using the SETTLE
>algorithm to keep the water rigid?

I don't know the answer here. It's probably been done by accident,
presumably even by me, but I don't have any real recollection. Things
would look obviously bad if one used a usual time step of 2 to 4 fs, since
a non-rigid water simulation would need a very short time step, probably on
the order of 0.5 fs.


AMBER mailing list
Received on Thu Feb 18 2021 - 18:30:01 PST
Custom Search