Rice Omic-scale Multilayer Biological Network Reconstruction And Tool Development

Understanding how metabolic reactions translate the genome of an organism into its phenotype is a grand challenge in biology, which ties gene function to phenotype through gene regulatory networks (GRNs), protein-protein interactions (PPIs) and molecular pathways. The traditional strategies for globally linking genetic variation to phenotypic diversity have been outlined, such as quantitative trait locus (QTL) and genome-wide association studies (GWAS). However, QTL and GWAS are able to statistically connect genotypes to phenotypes, without any recourse to known molecular interactions, whereas a molecular mechanistic description ties gene function to phenotype, through GRNs. Integration of regulatory information of an organism at different levels is expected to provide a good way for mapping genotypes to phenotypes.In our study, we firstly built a draft metabolic model of rice based on the gene-enzyme-reaction associated information extracted from public available databases. An automated dead-end filling approach based on endosymbiosis (DEF) was proposed for accurate filling gaps of metabolic model. Consequently, the draft metabolic network was carefully curated by system biology and experimental methods, including charge calculation, reaction balancing, reactionreversibility prediction, gap filling and so on. Furthermore, omic-scaleprotein-protein interactions and gene regulations were integrated into thecurated metabolic model to complete the reconstruction of the rice omic-scalemultilayer biological network. In addition, a comprehensive and integrativeapproach for accurate plant subcellular localization prediction (PSI) wasdeveloped, to compartmentalize the omic-scale multilayer networks intodifferent subcellular locations. Finally, RiceNetDB, a comprehensive regulatorydatabase of rice, was built for systematically storing and retrieving the omic-scalemultilayer networks of rice in order to facilitate its biomolecular regulatorymechanism analysis and gene-metabolite mapping. A viewer as one of thefunctional modules in RiceNetDB was developed for gene-centric multilayernetwork visualization in three dimensions (3D). The omic-scale multilayerregulatory network obtained in this study will be crucial for exploring thebiomolecular regulatory mechanism and understanding the genotype-phenotyperelationship of rice.