Mixture Regression Analysis of *-Seq Data

• Activity Number: 72
• Type: Contributed
• Date/Time: Sunday, July 31, 2011 : 4:00 PM to 5:50 PM
• Sponsor: Biometrics Section
• Author(s): Naim Rashid*+ and Wei Sun and Joseph G. Ibrahim
• Companies: The University of North Carolina at Chapel Hill and The University of North Carolina at Chapel Hill and The University of North Carolina
• Keywords: Mixture Regression ; Discrete Time Series ; ChIP-seq ; Next Generation Sequencing ; Negative Binomial ; Penalized Regression

Abstract

Next generation sequencing technologies are now routinely utilized for genome-wide investigation of various genomic activities, from the study of histone modifications to the detection of transcription factor binding sites. We refer to these collectively as *-seq experiments, which include Chromatin Immunoprecipitation (ChIP-seq), DNase hypersensitive site mapping (DNase-seq), and Formaldehyde-Assisted Isolation of Regulatory Elements (FAIRE-seq), among others. A common goal of the analysis of *-seq data is to identify genomic loci with enriched sequence reads. We discuss a general mixture regression approach for the analysis of *-seq data, where arbitrary sets of covariates are introduced to augment the identification of enrichment loci. We also introduce a new penalization procedure to efficiently identify covariates within each mixture component, and discuss approaches to deal with locally correlated responses in the mixture regression setting.