“…In this regard, as successfully applied to other systems, , the combination of docking calculations against an ensemble of equilibrium structures of the enzyme to handle protein flexibility and MM/PBSA calculations to estimate solvation energies is critical for technical aspects of the adopted approach to minimize such drawbacks. It is worth emphasizing that MM/PBSA estimates ⟨ W *( r , ω )⟩ in Table overestimate binding energies as computed either from QM/MM-based free energy perturbation , or umbrella sampling for a variety of AChE inhibitors, including huperzine A, tacrine, galantimine, phenserine, rivastigmine, dichlorvos, soman, huprines, and Cordyceps derivatives. Correction of the estimated potential of mean force W *( r , ω ) by means of a scaling factor α in order to reproduce the experimental equilibrium dissociation constant is therefore another critical aspect of the applied methodology devised to minimize such error, allowing for proper comparison of free energy stabilities of basins 1, 2, and 3–indeed, scaled free-energies ⟨ W *( r , ω )⟩ in Table are comparable to independent binding estimates for donepezil in the range of one or two dozens of kcal per mol, as determined by means of metadynamics using more accurate force fields .…”