[gmx-users] Binding free energy of an ion from alchemical transformation

[Wojciech Kopeć]

Dear Gromacs users,

I'm thinking about performing free energy calculations for a protein that
binds cations,  using alchemical transformation method. In the simplest
case I'd transform an ion (one atom) into a naked particle; however in this
case I'd get a non-zero overall charge of the system during the charge
decoupling. I'm aware of some corrections for that and also that Gromacs
probably neutralises the charge with PME (?). Another approach I've been
thinking of is to simultaneously couple a naked particle somewhere in the
bulk with the environment, switching on the charge in a same way as the
charge of the ion is being switched off. In this way, the charge would be 0
all the time, and at the end of the simulation would yield the protein with
an empty binding site and the ion somewhere in the bulk. Are there any
obvious pitfalls with this approach?

Thank you,
Wojciech