Large-scale Gene Association Network Inference and Functional Module Discovery in Plants

Accurately modelling and reconstructing genome-scale biological networks that play critical roles in many fundamental functions in plants, such as signaling transduction, metabolism, and gene expression regulation, is essential toward the understanding of gene function, interaction, and cellular behavior at the genome level. Yet, reverse-engineering of genome-scale biological networks and subsequently discovering functional modules remain a very computationally challenging task, which generally requires effective statistical approaches, innovative computational algorithms and efficient algorithm implementations. To infer, modeling and analyze complex heterogeneous biological networks, we developed a series of innovative bioinformatics methods and systems.

Network inference: We present 1) GPLEXUS (http://plantgrn.noble.org/GPLEXUS/) and 2) DeGNServer (http://plantgrn.noble.org/DeGNServer/), both of which integrate a series of innovative algorithms that have been implemented using parallel-computing techniques to accurately reconstruct and analyze genome-wide plant gene-gene association networks from medium- to large-scale expression profile data. The GPLEXUS adopts an ultra-fast estimation for pairwise mutual information (MI) computing while the DeGNServer extends the widely used Context Likelihood of Relatedness (CLR) framework by the integration of six proven correlation or association analysis methods, e.g. Spearman rank correlation, Pearson correlation, Mutual-information, Maximum information coefficient, Kendall rank correlation, Thei-sen Estimator; and both servers implement the algorithms in parallel and distributed environment to enable genome-scale gene-gene association inference at extraordinary speed while maintaining high prediction accuracy.

Network modeling, integration and graph-search empowered network analysis: We present HRGRN (http://plantgrn.noble.org/hrgrn/), which is a “graph search”-empowered integrative bioinformatics platform to host, analyze and discover interactions among genes, proteins, compounds, and small RNAs involving in signaling transduction, metabolic, and gene regulatory networks in Arabidopsis thaliana.

Subnetwork/functional module discovery: Much of cells’ functions are organized as networking of interacting modules or sub-networks. We present a novel approach to mine functional modules in heterogonous biological networks using Multiplex PageRank method. Starting from several seed genes, we successfully identified an immune-related core function module involved in plant defensive response to pathogens in A. thaliana through integrative analysis of gene co-expression-based association networks or protein-protein interaction networks. Further gene set enrichment analysis and literature analysis on this core module validates that our approach is effective and highly promising to mine functional modules in large-scale heterogeneous biological networks.