Arquitetura computacional para simulação de redes metabólicas em larga escala

Abstract

The amount of biological data is growing every day. Many organisms had their metabolic networks characterized. These studies permit a better understanding of cell behavior, and can be used to aid rational drug design and prediction of its side e®ects. One of the main tools to explore metabolic networks is computer simulations, but the size and complexity of networks impacts on the simulation costs.

These systems have a stochastic behavior, they can take di®erent paths depending on environmental conditions. Metabolic network simulation using a stochastic model computation can be distributed, but they have a high computational cost. On the other hand, the grid technology is evolving and becaming an alternative to supercomputers on high performance computing. The study of complex metabolic networks can be done on acceptable time by the employment of grids. In this work we describe an implementation of a computational architeture to execute this kind of large scale simulation. This architeture explores charecter