explained in Sire rafe_section the approach use to compute absolute free energy of hydration (namely the double annihilation scheme)
Clarified the Sire/SOMD results section. In particular, I gave a better explanation of the unperturbed hydrogen bond constraint, the bond-stretching error term and updated the table according with the current nomenclature
I kept some paragraphs where I am talking about mean unsigned error, with additional comments, since this may be useful in future (?)
updated SI for Sire with new tables showing results
In this revision I have :