Nonparametric Methods for Identifying Differential Binding Regions with ChIP-Seq Data

• Activity Number: 303
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Qian Wu*+ and Kyoung-Jae Won and Hongzhe Li
• Companies: University of Pennsylvania and University of Pennsylvania and University of Pennsylvania
• Keywords: ChIP-Seq ; Kernel Smoothing ; Normalization ; Nonparametric Testing

Abstract

ChIP-Seq provides a powerful method for detecting binding sites of DNA-associated proteins, e.g. transcription factors (TFs) and histone modification marks(HMs).Previous research has focused on developing peak-calling procedures for TFs. However, these procedures have difficulty when applied to HMs. It is also important to identify genes with differential binding (DB) regions between two conditions, such as different cellular states. Parametric methods based on Poisson/Negative Binomial distribution have been proposed to address this problem and most require biological replications. Many ChIP-Seq data usually have a few or even no replicates. We propose a novel nonparametric method to identify the DB regions that can be applied to the ChIP-Seq data of TF or HM, even without replicates. Our method is based on nonparametric hypothesis testing and kernel smoothing. We prove the method using a ChIP-Seq data of human adipose stromal cells and it detects nearly 20% of genes with differential binding of HM mark H3K27ac in gene promoter regions. The test statistics also correlate with the gene expression changes well, indicating the identified DB regions are indeed biologically meaningful.