Co-expression Gene Network Analysis and Functional Module Identification in Bamboo Growth and Development

Xuelian Ma,Hengyu Yan,Hansheng Zhao,Wenying Xu,Qi You,Zhimin Gao,Zhen Su

doi:10.3389/fgene.2018.00574

Abstract

Bamboo is one of the fastest-growing non-timber forest plants. Moso bamboo (Phyllostachys edulis) is the most economically valuable bamboo in Asia, especially in China. With the release of the whole-genome sequence of moso bamboo, there are increasing demands for refined annotation of bamboo genes. Recently, large amounts of bamboo transcriptome data have become available, including data on the multiple growth stages of tissues. It is now feasible for us to construct co-expression networks to improve bamboo gene annotation and reveal the relationships between gene expression and growth traits. We integrated the genome sequence of moso bamboo and 78 transcriptome data sets to build genome-wide global and conditional co-expression networks. We overlaid the gene expression results onto the network with multiple dimensions (different development stages). Through combining the co-expression network, module classification and function enrichment tools, we identified 1,896 functional modules related to bamboo development, which covered functions such as photosynthesis, hormone biosynthesis, signal transduction, and secondary cell wall biosynthesis. Furthermore, an online database (http://bioinformatics.cau.edu.cn/bamboo) was built for searching the moso bamboo co-expression network and module enrichment analysis. Our database also includes cis-element analysis, gene set enrichment analysis, and other tools. In summary, we integrated public and in-house bamboo transcriptome data sets and carried out co-expression network analysis and functional module identification. Through data mining, we have yielded some novel insights into the regulation of growth and development. Our established online database might be convenient for the bamboo research community to identify functional genes or modules with important traits.

Highlights

Bamboo, an important fast-growing non-timber forest plant worldwide, has been an essential forest resource with an annual trade value of >2.5 billion US dollars, and approximately 2.5 billion people depend on it economically (Peng et al, 2013a,b; Zhao et al, 2017)
We integrated 78 transcriptome data sets for moso bamboo (Phyllostachys edulis), which can be divided into two parts according to the data source: 52 public data sets from NCBI and 26 in-house data sets from International Center for Bamboo and Rattan (ICBR) (Table 1)
Through Gene Ontology (GO) enrichment analysis of all genes from this network by using agriGO (Du et al, 2010; Tian et al, 2017) (Figure 1D), the results showed that these co-expressed genes were strongly associated with the GO terms of photosynthesis and light harvesting, light reaction, and generation of precursor metabolites and energy, which matched the previous findings that the primary function of light-harvesting complex (LHC) protein was the absorption of light through chlorophyll excitation and transfer of absorbed energy to photochemical reaction centers (Dolganov et al, 1995; Li et al, 2000; Montané and Kloppstech, 2000; Zhao et al, 2016)

Summary

Introduction

An important fast-growing non-timber forest plant worldwide, has been an essential forest resource with an annual trade value of >2.5 billion US dollars, and approximately 2.5 billion people depend on it economically (Peng et al, 2013a,b; Zhao et al, 2017). Through data mining tools and algorithms that describe complex co-expression patterns of multiple genes in a pairwise fashion, global co-expression network analyses consider all samples (multiple data sources with independence) together and establish connections between genes based on the collective information available (Bassel et al, 2011) Compared with such a network, the conditional coexpression network aims to enhance our understanding of gene function from a portion of transcriptome data sets that have much in common, such as having the same source and a similar acquisition of raw materials and inferring gene transcriptional regulatory mechanisms in developmental processes based on a series of selected associated samples. Co-expression networks with expression views can be used to associate genes of unknown function with biological processes, to discern gene transcriptional regulatory mechanisms in vivo and to prioritize candidate regulatory genes or modules of vital traits

Methods

Results

Discussion

Conclusion