(184t) Accurate Calculation of Log P_OW Values From Atomistic Computer Simulations | AIChE

(184t) Accurate Calculation of Log P_OW Values From Atomistic Computer Simulations

Authors 

Köddermann, T. - Presenter, Fraunhofer Institute for Algorithms and Scientific Computing (SCAI)
Reith, D. - Presenter, Fraunhofer Institute for Algorithms and Scientific Computing (SCAI)
Arnold, A. W. - Presenter, Fraunhofer Institute for Algorithms and Scientific Computing (SCAI)


The log POW value of a substance measures its solubility in octanol compared to water. This can be seen as a simply model for its solubility in biological membranes, which ultimately is a rough estimate for toxicity. If a substance is hardly soluble in octanol, it is practically impossible for it to enter (human) cells, and therefore is less likely to be toxic. On the other hand, for novel drugs it might be important to penetrate the cell through the membrane, or even integrate into it.

Being able to determine log POW values a priori from computer simulations would therefore be an important step towards virtual material design. At present, log POW values are mainly determined by heuristic approaches based on the chemical structure. We present a method to reliably calculate log POW values from atomistic computer simulations. Our method is based on the calculation of solvation energies in water and octanol using thermodynamic integration. First results demonstrate that this generic approach is able to predict log POW values for alcohols significantly better than the classical heuristic methods. In addition, we show that our method can also be applied to substances with concentration-depend log POW values, such as ionic liquids.