This is the proposed framework for the theory section we discussed. It uses established statistical mechanics theories of ideal dilute solutions and the theory of binding from Gilson et al. '97. It covers solvation and partitioning on the same footing as binding.

I believe that all alchemical protocols discussed (ABFE, RBFE, DDM, ATM, SepTop, etc.) can be derived from Eq. 23 (Gilson et al.'s expression of the binding constant). I also believe that topics such as multiple binding poses, chemical heterogeneity, the use of restraints, etc. can all be framed in this context.

The math in the section on reversible binding "Simulating binding events is computationally expensive" is optional. I put it there to illustrate the derivation of Eq. 4 in the v 1.0 version from Eq. 23 of the new version.

Comments/suggestions/opinions are welcome! This is only a proposal.