Data Mining em aplicações de Desenho racional de Fármacos

Abstract

The development of a new drug is a long process, very expensive and raises some ethical questions.The creation of a new medicine goes through four phases:drug discovery, pre-clinical development, clinical development and regulatory approval.The use of Data Mining (DM) techniques in pre-clinical development phase can accelarate and cheapen the creation of a new drug.This phase is divided in three subphases: pharmacology, toxicological ordination and formulation, in wich will be used the DM techniques to optimize the process.The classification and regression algorithms, as well as the techniques of a data mining to use will use repositories compouds already tested by pharmacological phases and taxicological ordination, in based of ADMET tests, and the rules defined as useful, by one skilled, to the design process of new drugs. In that case the techniques of data mining prediction can use known results of previous tests and make a prediction about the taxicological properties and ADME.In short, it is intended to develop an interface between an expert and data mining tools that allow interaction interface with the repositories of molecules, build models using the Data Mining, view the models in a format used by biochemists and allow interaction to modify the built models.