Chem. Pharm. Bull. 55(4) 669—674 (2007)

tant molecular property, playing a large role in the behavior of compounds in many areas of interest. Given the importance of solubility, a means of prediction based solely on molecular structure should prove a useful tool, as many compounds exist for which the solubility simply is not available. The solubility of chemicals and drugs in the water phase has an essential influence on the extent of their absorption and transport in a body. That is why solubility is considered to be a very important parameter in current ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) research. Water solubility plays a key role in areas such as drug dosage, anesthesiology, corrosion of metals, transport fate of pollutants in terrestrial, aquatic and atmospheric ecosystems, deposition of minerals and composition of ground waters, and availability of oxygen and other gases in life support systems. The widespread relevance of water solubility data to many branches and disciplines of science, medicine, technology, and engineering has led to the development of several models to predict water solubility. Hence, it was deemed advantageous to develop a model to predict water solubility using only theoretically derived descriptors. Comparing with the time-consuming experimental procedures to determine aqueous solubility directly, reliable computational methods to predict aqueous solubility are more popular in today’s research. There are some reports about the applications of QSPR approaches to predict the aqueous solubility of organic compounds. In our previous papers, we reported on the application of QSPR techniques in the development of a new, simplified approach to prediction of compounds properties. Several articles have published with MLR models for the prediction of aqueous solubility. In a QSPR study, a mathematical model is developed which relates the structure of a set of compounds to a physical property such as aqueous solubility. In a QSPR study is that there is some sort of relationship between the physical property of interest and structural descriptors. These descriptors are numerical representations of structural features of molecules that attempt to encode important information that causes structurally different compounds to have different physical property values. Even though the descriptors used to build a QSPR model can be empirical, it is generally more useful to use descriptors derived mathematically from the 3D molecular structure, since this allow any relationship so derived to be extended to the prediction of the property for unavailable compounds. In this work a QSPR study is performed, to develop model that relate the structures of a heterogeneous group of 150 drug-like compounds to their aqueous solubility. The stepwise MLR was used to select the most informative descriptors from the calculated descriptors by Molecular Modeling Pro Plus software. The selected descriptors were used to develop a MLR model for predicting the solubility for 40 drug compounds in water at 25 °C. The aim of this work was to investigate molecular descriptors important in determining aqueous solubility.