Strategy for multivariate Identification of diferentially expressed genes in microarray data

Abstract. Microarray technology has become one of the most important tools in understanding genetic expression in biological processes. As microarrays contain measurements of thousands of genes' expression levels across multiple conditions, identification of differentially expressed genes will necessarily involve data mining or large scale multiple testing procedures. To the date, advances in this regard have either been multivariate but descriptive, or inferential but univariate. In this work, we present a new multivariate inferential analysis method for detecting differentially expressed genes in microarray data. It estimates the positive false discovery rate (pFDR) using artificial components close to the data's principal components, but with an exact interpretation in terms of differential gene expression. Our method works best under very common assumptions and gives way to a new understanding of genetic differential expression in microarray data. We provide a methodology to analyse time course microarray experiments and some guidelines for assessing whether the required assumptions hold. We illustrate our method on two publicly available microarray data sets.

• Autores: Acosta Rivera, Juan Pablo
• Fecha de publicación: 2015
• Institución: Universidad Nacional de Colombia
• Repositorio: Universidad Nacional de Colombia
• Idioma: spa
• OAI Identifier: oai:repositorio.unal.edu.co:unal/54111
• Acceso en línea: https://repositorio.unal.edu.co/handle/unal/54111; http://bdigital.unal.edu.co/48941/
• Palabra clave: 51 Matemáticas / Mathematics; 57 Ciencias de la vida; Biología / Life sciences; biology; Microarrays; False Discovery Rate; Principal Components Analysis; Bootstrap; Microarreglos de ADN; Tasa de falsos positivos; Análisis en componentes principales,
• Rights: openAccess
• License: Atribución-NoComercial 4.0 Internacional