Pmx ABFE ligand only simulation set-up


I’m trying to set up ABFE calculations for a guest-host system in water following the pmx abfe functionality. The complex works, although has some trouble minimising. I let it generate the guest-host constraints for the FEP step automatically.
I was then hoping to run the second arm of the calculation with guest molecule (Glycine anhydride) only but I was not sure how to impose the restraints then, and could not find any pmx documentation on the matter.
Clearly using the original set does not work but then am I free to choose any heavy atoms?
I tried running a calculation but it didn’t seem to restrain the position as found in .xtc and the simulation stops prematurely without a clear error (Segmentation fault (core dumped)).

Thanks in advance!
Katarzyna Zator

It is not clear why you would want restraints in the guest-only leg of the thermodynamic cycle. One usually only needs to remove relative rotation+translation of the guest with respect to host for the host-guest branch of the cycle.


Ah, I was under the impression they were both done explicitly as diagrams are usually drawn so.
How can I calculate the rot+trans contributions then?

Analytical expression for the removal of rot+trans degrees of freedom is derived here:

pmx calculates it when setting up the abfe and prints out to a separate file.