A0680
Title: Discovering dependence structure of transcription factors based on a nonhomogeneous Poisson process model
Authors: Xiaowei Wu - Virginia Tech (United States) [presenting]
Hongxiao Zhu - Virginia Tech (United States)
Abstract: Transcription factors (TFs) are proteins that bind to DNA sequences, playing the central role of regulating gene expressions. Discovering the dependence structure among TFs is a critical problem in statistical genetics, and it advances the understanding of the underlying epigenetic regulatory mechanism. Based on a nonhomogeneous Poisson process (NHPP) model, a two-step approach is developed for detecting pairwise TF relations, whether they bind to DNA sequences cooperatively or independently. The key of the proposed method lies in the goodness-of-fit test of the NHPP model to recurrent TF binding events, which is an important but often neglected issue in modeling complex data arising from real-world problems. Simulation studies show that the proposed method provides a powerful test for interactions between TFs. The method is applied to analyze ChIP-seq data of 14 TFs collected in recent mouse embryonic stem cell research. Findings provide new insights into the orchestration of TFs in gene regulation processes.
