A Database of Pharmacological Properties For Accurate ADME-Tox Predictions

SANCHEZ J.; BESSONE S; VAZQUEZ G.E; PARDILLO FONTDEVILA E; BRIGNOLE N.B

Rio das Pedras, Brasil

Congreso; ENPROMER 2005; 2005

The prediction of Absorption, Distribution, Metabolism, Excretion and Toxicity (ADME-Tox) properties is crucial for drug design because most of the candidates are rejected for having failed some ADME-Tox test. Since a huge amount of candidate compounds is normally explored, and the determination of ADME-Tox properties by experimental methods is both expensive and time-consuming, there is a need for computational procedures that yield trustworthy predictions based exclusively on the chemical structure of the prospective drug. This software can be employed to choose the most promising candidates, thus limiting the number of potential drugs that undergo the costly in-vitro and in-vivo tests. The quality of the in-silico results strongly depends on both the accuracy of the data and the relevance of the descriptors chosen to model each property. The former implies the need for a reliable database, while the latter involves the design of adequate functionalities and sound criteria for the selection of the set of existing compounds employed to find the correlation coefficients. In this work the development of an ad-hoc database of medicinal compounds for the support and retrieval of information concerning molecular descriptors is presented. A simple example of the application of the database to logP prediction is also described. This property is closely related to the distribution of a drug through the organism, as well as to its interaction with the receptors.