Statistical Analysis of Dynamic Transcriptional Regulatory Network Structure

Abstract

Here, we present a detailed method for generating a dynamic transcriptional regulatory network from large-scale chromatin immunoprecipitation data, and functional analysis of participating factors through the identification and characterization of significantly overrepresented multi-input motifs in the network. This is done by visualizing interactive data using a network analysis tool, such as Cytoscape, clustering DNA targets of the transcription factors based on their network topologies, and statistically analyzing each cluster based on its size and properties of its members. These analyses yield testable predictions about the conditional and cooperative functions of the factors. This is a versatile approach that allows the visualization of network architecture on a genome-wide level and is applicable to understanding combinatorial control mechanisms of DNA-binding regulators that conditionally cooperate in a wide variety of biological models.