Towards calculation of hydration free energy
Volodymyr Sergiievskyi
Hydration Free Energy (HFE) is one of the fundamental quantities in Physical Chemistry of solutions. It allows to predict many properties of compounds, such as solubility, pKa, constants of complex formation. Application of HFE prediction methods are found in pharmacology. For screening large drug-candidate datasets the drug-design industry needs fast method for prediction HFE with the acceptable method accuracy 1-2 kcal/mol. Promising tool for such task is the integral equation theory, in particular Reference interaction site model (RISM) [1]. It is a compromise between slow MD or MC simulations and fast, but too-much simplified polarizable continuum models. Although RISM is much faster than MD/MC simulations, the standard algorithms of solving the RISM equation are not fast enough for effective screening of large molecule sets.
In the work [2] it was shown, that the multi-grid method, applied to the equation for the monoatomic spherical particles, reduces the computational cost of standard algorithm up to several dozen times. Now we applied the multi-grid technique to a more common case: RISM equations for polyatomic solvent. It was shown, that RISM multi-grid solver is stable for all investigated physically-reasonable systems and give more than 30 times speed-up of the standard algorithm. With the new algorithm the time to calculate Hydration Free Energy of the typical drug-like molecule ( 20 atoms) is about 15 sec instead of several minutes with the standard algorithm. Recently performance of the new multi-grid RISM solver for prediction Hydration Free Energy of drug-like bioactive compounds was studied.[3] It was shown, that RISM Hydration Free Energy calculations with structures and partial charges obtained by the QM methods give a good correlation with experiment, and after simple atom-type parametrization are able to give RMSD about 1.2 kcal/mol. This shows the perspective to use the efficient RISM-based method for calculating HFE in industrial applications.
References
- Chandler, D. and Andersen, H. C. J. Chem. Phys. 57(5):1930-1937, (1972)
- Fedorov, M. V. and Hackbusch, W. Preprint, 88, Max-Planck-Institut fuer Mathematik in den Naturwissenschaften (2008)
- David S. Palmer, Volodymyr P. Sergiievskyi, Frank Jensen, Maxim V. Fedorov, accepted for publication in J. Chem. Phys., Manuscript No.: A10.04.0425R