Rigidwater not working in plugin

[Andrea Zonca]

Hello,

in our plugin the rigidwater flag does not work correctly.
We have a simple test case of 3 water molecules, see https://github.com/sqamara/mbpol_debug/ ... er/test.py

If we setup one of the OpenMM water models, like TIP5P, everything works correctly, see a plot of OH distances for the first molecule for 100 steps at .2 fs for rigidwater true and false:

TIP5P distances

Instead, with our plugin, the distances drift in time:

MBPol distances

We checked in `forcefield.py` and it looks like the `isConstrained` is correctly set to `true`.
However in the simulation the water is not rigid.

We think there is something wrong in the plugin code, however we cannot figure out where to look,
would you have any hint?

thanks!

[Peter Eastman]

Water molecules are identified by the residue name in the Topology: HOH. Did you name them something different from that?

Peter

[Andrea Zonca]

Thanks, no, I use HOH as well, see the xml:

https://github.com/paesanilab/mbpol_ope ... /mbpol.xml