Dynamic gene regulatory network reconstruction and analysis based on clinical transcriptomic data of colorectal cancer.

Inferring dynamic regulatory networks that rewire at different stages is a reasonable way to understand the mechanisms underlying cancer development. In this study, we reconstruct the stage-specific gene regulatory networks (GRNs) for colorectal cancer to understand dynamic changes of gene regulations along different disease stages. We combined multiple sets of clinical transcriptomic data of colorectal cancer patients and employed a supervised approach to select initial gene set for network construction. We then developed a dynamical system-based optimization method to infer dynamic GRNs by incorporating mutual information-based network sparsification and a dynamic cascade technique into an ordinary differential equations model. Dynamic GRNs at four different stages of colorectal cancer were reconstructed and analyzed. Several important genes were revealed based on the rewiring of the reconstructed GRNs. Our study demonstrated that reconstructing dynamic GRNs based on clinical transcriptomic profiling allows us to detect the dynamic trend of gene regulation as well as reveal critical genes for cancer development which may be important candidates of master regulators for further experimental test.

Exo-miRExplorer: A Comprehensive Resource for Exploring and Comparatively Analyzing Exogenous MicroRNAs.

MicroRNAs (miRNAs) are small regulatory RNAs that play important roles in animals, plants, and viruses. Deep-sequencing technology has been widely adopted in miRNA investigations. However, it is still a big mysterious why nearly all sequencing data contain miRNA sequences from exogenous species, called exo-miRNAs. In this study, we developed a novel platform, exo-miRExplorer, for mining and identifying exo-miRNAs from high-throughput small RNA sequencing experiments which originated from tissues and cell lines of multiple organisms. Thousands of exo-miRNAs are characterized with their expression abundance, the RNA families, original organisms and the sequencing platforms presented in exo-miRExplorer. Subsequently, we used exo-miRExplorer to perform further analysis. Comparative analysis of the exo-miRNAs between different sequencing datasets revealed significant correlation of exo-miRNAs between experiments in the same study. The plant-derived exo-miRNAs analysis provided robust evidence for non-diet source of exo-miRNAs. Virus-derived exo-miRNA analysis showed that pathogen RNAs could transfer to host cells and exist in deep-sequencing result at abundance level. In conclusion, exo-miRExplorer provides users with an integrative resource to facilitate detection and analysis of exo-miRNAs. exo-miRExplorer is available at the following URL: http://rna.sysu.edu.cn/exomiRDB/.