文摘
Continuum-solvent models (CSMs) have successfully predicted many quantities, including the solvation-free energies (螖G) of small molecules, but they have not consistently succeeded at reproducing experimental binding free energies (螖螖G), especially for protein鈥損rotein complexes. Several CSMs break 螖G into the free energy (螖Gb>vdwb>) of inserting an uncharged molecule into solution and the free energy (螖Gb>elb>) gained from charging. Some further divide 螖Gb>vdwb> into the free energy (螖Gb>repb>) of inserting a nearly hard cavity into solution and the free energy (螖Gb>attb>) gained from turning on dispersive interactions between the solute and solvent. We show that for 9 protein鈥損rotein complexes neither 螖Gb>repb> nor 螖Gb>vdwb> was linear in the solvent-accessible area A, as assumed in many CSMs, and the corresponding components of 螖螖G were not linear in changes in A. We show that linear response theory (LRT) yielded good estimates of 螖Gb>attb> and 螖螖Gb>attb>, but estimates of 螖螖Gb>attb> obtained from either the initial or final configurations of the solvent were not consistent with those from LRT. The LRT estimates of 螖Gb>elb> differed by more than 100 kcal/mol from the explicit solvent model鈥檚 (ESM鈥檚) predictions, and its estimates of the corresponding component (螖螖Gb>elb>) of 螖螖G differed by more than 10 kcal/mol. Finally, the Poisson鈥揃oltzmann equation produced estimates of 螖Gb>elb> that were correlated with those from the ESM, but its estimates of 螖螖Gb>elb> were much less so. These findings may help explain why many CSMs have not been consistently successful at predicting 螖螖G for many complexes, including protein鈥損rotein complexes.