“…Geometry does not completely explain the observed CPE distributions though. In general, across all simulations, there is not a smooth progression from a relatively higher to relatively lower CPE as the surfaces increase, as one would expect if geometry alone explained the data . Instead, for all of the OPLS simulations (OPLS-Std, OPLS-RESP, and OPLS-CM5), we observe a “back-and-forth” behavior where surfaces 1 and 2 have higher CPE, surface 5 has low CPE, and surfaces 3 and 4 oscillate between high and low CPE.…”