Publication: Contrastes de Hipótesis Múltiples Bajo Dependencia con Aplicación a los Microarrays: Una Aproximación Bayesiana
Loading...
Files
Official URL
Full text at PDC
Publication Date
2021-05-07
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad Complutense de Madrid
Citations
Exportar
Abstract
Muchos experimentos requieren contrastar simultáneamente un elevado número de hipótesis. Un ejemplo son los experimentos con microarrays de ADN en el campo de la genómica, donde es habitual analizar simultáneamente un gran número de genes con la finalidad de identificar cuáles de ellos se expresan de manera diferencial bajo dos condiciones experimentales. Uno de los problemas que se presentan en este contexto, además de la necesidad de contrastar simultáneamente un elevado número de hipótesis, uno para cada gen, es cómo modelar la dependencia que suele existir en el nivel de expresión entre los genes. La literatura al respecto es extensa, sin embargo, los procedimientos propuestos, en gran parte desde un punto de vista frecuentista y bajo el supuesto de independencia, no resuelven, en general, todos los problemas planteados anteriormente. El objetivo principal de esta tesis es la identificación de genes con expresión diferencial bajo dos condiciones de tratamiento distintas e independientes y bajo dependencia en el nivel de expresión de los genes. Para ello, se propone un procedimiento bayesiano en el que la dependencia se modela mediante funciones cópulas...
Many experiments require the simultaneous testing of a large number of hypotheses. An example of this, are the experiments with DNA microarrays in the field of genomics, where it is usual to simultaneously analyze a large number of genes in order to identify which of them are differentially expressed under two experimental conditions. One of the problems that arise in this context, in addition to the need to simultaneously test a large number of hypotheses, one for each gene, is how to model the dependence that usually exists in the level of expression between genes. The literature on the matter is extensive. However, the proposed procedures, most from a frequentist point of view and under the assumption of independence, do not usually solve the problems raised above.The main objective of this thesis is the identification of genes with differential expression in two different and independent treatment conditions by considering the dependence on the expression level of the genes. To achieve this goal, a Bayesian procedure is proposed in which the dependency is modelled using copula functions...
Many experiments require the simultaneous testing of a large number of hypotheses. An example of this, are the experiments with DNA microarrays in the field of genomics, where it is usual to simultaneously analyze a large number of genes in order to identify which of them are differentially expressed under two experimental conditions. One of the problems that arise in this context, in addition to the need to simultaneously test a large number of hypotheses, one for each gene, is how to model the dependence that usually exists in the level of expression between genes. The literature on the matter is extensive. However, the proposed procedures, most from a frequentist point of view and under the assumption of independence, do not usually solve the problems raised above.The main objective of this thesis is the identification of genes with differential expression in two different and independent treatment conditions by considering the dependence on the expression level of the genes. To achieve this goal, a Bayesian procedure is proposed in which the dependency is modelled using copula functions...
Description
Tesis inédita de la Universidad Complutense de Madrid, Facultad de Ciencias Matemáticas, leída el 18-12-2020