Heuristic lipophilicity potential for computer-aided rational drug design:Optimizations of screening functions and parameters

lIn this research we test and compare three possible atom-based screening f unctions used in the heuristic molecular lipophilicity potential (HMLP). Sc reening function 1 is a power distance-dependent function, b(i)/ // R-i - r //(gamma), screening function 2 is an exponential distance-dependent funct ion, b(i) exp(-// R-i - r // /d(0)), and screening function 3 is a weighted distance-dependent function, sign(b(i)) exp[-xi(// R-i - r // / / b(i) /)] . For every screening function, the parameters (gamma, d(0), and xi) are op timized using 31 common organic molecules of 4 types of compounds: aliphati c alcohols, aliphatic carboxylic acids, aliphatic amines, and aliphatic alk anes. The results of calculations show that screening function 3 cannot giv e chemically reasonable results, however, both the power screening function and the exponential screening function give chemically satisfactory result s. There are two notable differences between screening functions 1 and 2. F irst, the exponential screening function has larger values in the short dis tance than the power screening function, therefore more influence from the nearest neighbors is involved using screening function 2 than screening fun ction 1. Second, the power screening function has larger values in the long distance than the exponential screening function, therefore screening func tion 1 is effected by atoms at long distance more than screening function 2 . For screening function 1, the suitable range of parameter gamma is 1.0 < gamma < 3.0, gamma = 2.3 is recommended, and gamma = 2.0 is the nearest int egral value. For screening function 2, the suitable range of parameter do i s 1.5 < d(0) < 3.0, and d(0) = 2.0 is recommended. HMLP developed in this r esearch provides a potential tool for computer-aided three-dimensional drug design.