Predicting Functional Targets of Transcription Factors by Coupling CHiP-Seq with Geo

• Activity Number: 586
• Type: Invited
• Date/Time: Thursday, August 2, 2012 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Hongkai Ji*+ and George Wu
• Companies: The Johns Hopkins University and The Johns Hopkins University
• Keywords: ChIP-seq ; Gene expression ; Next-generation sequencing ; Bayesian model ; Data integration

Abstract

Chromatin immunoprecipitation coupled with massively parallel sequencing (ChIP-seq) is increasingly used to study the role of transcription factors (TFs) in gene regulation. A ChIP-seq experiments can identify thousands of target genes, but typically only a fraction of the targets respond to TF binding, while the others are likely to represent non-functional binding. For this reason, ChIP experiments are often coupled with gene expression profiling to more accurately define functional targets. However, we found that a large proportion of studies did not provide corresponding gene expression data. Based on the previously observed correlation between TF and target genes, we hypothesized that publicly available gene expression data can be used to more effectively identify functional ChIP target genes by combining the likelihood of TF binding with the correlation between gene and TF expression. We developed a model to integrate ChIP-seq with data in Gene Expression Omnibus to accomplish this goal. Our test show that this new method can significantly improve functional target prediction.