Estudos das relações estrutura-toxicidade de agrotóxicos

Abstract

Computational methods such as electrostatic and structural similarity of compounds were used to determine a similarity pattern for pesticides present in the database of the Medicinal Chemistry Group IQSC/USP. A collection consisting of herbicides and insecticides allowed and banned, according to the World Health Organization was used in this study. Several chemometric analyzes were performed to determine the recognition pattern processes in the classification of substances into their chemical classes that were associated with levels of 61-86% of correct prediction. Three case studies were carried out to validate the established hypotheses which showed good results in all cases. The disclosure of chemical space associated with the biological space using robust methods of analysis led to the establishment of rules for the selection and identification of patterns that can now be used as the foundation for the design of new herbicides and insecticides.