ALGORITMO MEMÉTICO BASADO EN GPU CON DIVERSIDAD Y ESTRUCTURA PARA LA INFERENCIA DE MODELOS DE REDES REGULADORAS DE GENES

FLORES OLIVAS, ANGEL OMAR (2017) ALGORITMO MEMÉTICO BASADO EN GPU CON DIVERSIDAD Y ESTRUCTURA PARA LA INFERENCIA DE MODELOS DE REDES REGULADORAS DE GENES. Maestría thesis, UNIVERSIDAD AUTONOMA DE CHIHUAHUA.

Vista Previa

Text Tesis.pdf Download (3877Kb) | Vista Previa

Resumen

Gene regulatory networks inference is defined as the analysis of expression levels of the genes on a network to determine the effects of interactions among them. This analysis is possible due to the specific properties of this network type. The knowledge of this kind of relationships has significant importance and uses, varying from behavioral prediction of a medical treatment in a patient or a personalized medicine, to the smart design of organisms, such as the design of bacterium for material decomposition process. Nowadays, finding Gene regulatory networks is a complicated process that requires advanced instrumentation, as well as complicated and expensive procedures. However, some approaches are used to obtain an approximation. Although in Computer science exist several way to tackle this problem, Memetic algorithms using differential evolution combined with a PSO algorithm as local search, are one of the promising techniques to offer quality solutions. Using this approach, the inference problem is modeled through differential equations known as S systems. In this thesis, an algorithm that integrates a group of the main approaches, from the state of the art, is used to speed up and optimize the functions of a memetic algorithm for gene regulatory networks inference.

Actions (login required)

Ver Objeto