Step-by-step construction of gene co-expression networks from high-throughput RNA sequencing data

The rapid increase in the availability of transcriptomics data generated by RNA sequencing represents both a challenge and an opportunity for biologists without bioinformatics training. The challenge is handling, integrating, and interpreting these data sets. The opportunity is to use this information to generate testable hypothesis to understand molecular mechanisms controlling gene expression and biological processes (Fig. 1). A successful strategy to generate tractable hypotheses from transcriptomics data has been to build undirected network graphs based on patterns of gene co-expression. Many examples of new hypothesis derived from network analyses can be found in the literature, spanning different organisms including plants and specific fields such as root developmental biology. In order to make the process of constructing a gene co-expression network more accessible to biologists, here researchers from the Pontifical Catholic University of Chile provide step-by-step instructions using published RNA-seq experimental data obtained from a public database. Similar strategies have been used in previous studies to advance root developmental biology. This guide includes basic instructions for the operation of widely used open source platforms such as Bio-Linux, R, and Cytoscape. Even though the data they used in this example was obtained from Arabidopsis thaliana, the workflow developed in this guide can be easily adapted to work with RNA-seq data from any organism.

Pipeline overview for data analysis

Light gray square boxes represent data files or data objects. Dark gray boxes represent tools or procedures. Black arrows correspond to direct steps. Light gray background shadows comprise sections in the chapter. (a) Data processing and (b) co-expression network generation