flexible water models

[Daniel Konstantinovsky]

Hi!
Does OpenMM have (or plan to have) an implementation of a flexible water model like SPC/Fw that is compatible with AMBER, etc. force fields for biomolecules? I am considering doing some flexible water MD but I'm not sure I want to switch to something as expensive as AMOEBA right away.
Thank you!

[Peter Eastman]

If you specify "constraints=None, rigidWater=False" in your call to createSystem(), you'll get flexible water. I just checked a couple of the Amber14 water models and verified that they include bond and angle terms, so presumably they were designed to support that.

[Daniel Konstantinovsky]

I see. But doing that wouldn't result in any of the "recognized" flexible models without changing charge parameters, etc. in the SPC/E or TIP?P model. Simply letting go of the rigidity seems like a quick-and-dirty way to get flexible water, but I anticipate problems justifying the model...What do you think?

[Peter Eastman]

You could create your own XML file defining the water model you want. Be aware though that Amber parametrizes the ions for a specific water model, so different water models include different ion parameters. Anything you do will be hard to justify in one way or another. Amber was parametrized with rigid water.

[Daniel Konstantinovsky]

Oh ok, I understand. It seems like AMOEBA is a better choice for protein containing systems with flexible water, but AMBER + flexible water may be useful for trial runs, etc.