Multiple Testing for Differential Expression Using RNA-Seq Data

• Activity Number: 629
• Type: Topic Contributed
• Date/Time: Thursday, August 8, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Mental Health Statistics Section
• Author(s): Dan Nettleton*+
• Companies: Iowa State University
• Keywords: False Discovery Rate ; Multiple Testing ; Simultaneous Inference ; RNA-Seq ; microarray ; generalized linear model

Abstract

Next Generation Sequencing of RNA (RNA-Seq) yields discrete counts associated with transcript abundance for each of thousands of transcripts and a relatively small number of samples. Scientists using RNA-Seq technology often wish to test for differential expression between different types of samples. This problem can be translated into statistical terms as a test of parameter equality in each of thousands of generalized linear models. We will examine the unique challenges that arise in this context when attempting to control or estimate false discovery rate or predict false discovery proportion. Methods for hierarchical modeling that allow sharing of information across genes and methods for addressing inter-gene dependence will be discussed.