Multivariate Profiling Approach to Inference on Differential Expression in RNA-Seq Data with Small Sample Size

• Activity Number: 571
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Sunghee Oh*+ and Mi-Ok Kim and Seongho Song
• Companies: Cincinnati Children's Hospital Medical Center and Cincinnati Children's Hospital Medical Center and University of Cincinnati
• Keywords: RNA-Seq ; Differential expression ; Univariate multivariate profiling approach ; Monte carlo simulation study

Abstract

Identifying differential expression is one of the most widely used in down-stream statistical analyses of genomic data. We propose a multivariate profiling approach to the identification of differential expression in RNA-Seq data with small sample size. RNA-Seq is a recently developed approach to transcriptome profiling that uses deep-sequencing technologies. It is fast replacing microarray with superior sensitivity and precision, dynamic resolution, and isoform level quantification. Contrary to its gaining popularity, the technology is still expensive and often only a small number of samples are included in individual experiments. The proposed fully bayesian multivariate approach will overcome limitations rising from the small sample size issue by pooling information across genes. This idea of multivariate profiling is well justified by well-known facts that genes work collaboratively in a biological system as a network module and the functional connections are manifested as correlated expressions among genes. Many multivariate approaches have