Um estudo comparativo de algoritmos de agrupamento de dados para dados de docagem molecular

Abstract

In Bioinformatics, we deal with a big amount of biology-related data, trying to obtain, from them, valuable information. Rational Drug Design is one field of research in bioinformatics, which is developed from the interaction between two molecules, called Receptor and Ligand. Through molecular docking simulations, we try to find the best positioning of these two molecules, searching for the best possible binding between them. The binding quality is measured by the FEB - Free Energy of Binding. Through these simulations, the best ligands to a receptor are tested in vitro and a new drug may be created. A strategy to get these results regards considering both the receptor and the ligand as flexible structures. However, the computational cost increases when considering such a flexibility. Hence, the 3D-Tri algorithm tries to find promising conformations to further molecular dockings experiments from a decision-tree strategy based on clustering analysis. The objective of this work is to treat this large amount of data, as well as contribute to evolve the 3D-Tri algorithm, with the analysis of data clustering algorithms to a possible time reduction in the execution of future docking experiments.