Inferring the perturbed microRNA regulatory networks from gene expression data using a network propagation based...

Area of Science:

• Genomics
• Bioinformatics
• Systems Biology

Background:

• MicroRNAs (miRNAs) are small regulatory RNAs crucial for cellular processes.
• Identifying dys-regulated miRNAs and their targets is key to understanding regulatory networks.
• Existing computational methods often overlook secondary effects of miRNA perturbation on global networks.

Purpose of the Study:

• To develop a novel computational method for inferring perturbed miRNA regulatory networks.
• To identify key microRNAs and their target genes by integrating gene expression and global regulatory network information.
• To improve the accuracy of miRNA perturbation network inference.

Main Methods:

• A network propagation approach using random walk with restart on gene regulatory networks.
• Modeling the network effects of miRNA perturbation.
• Evaluating the correlation between network effects and gene differential expression using a forward searching strategy.

Main Results:

• The proposed method outperformed existing approaches in identifying experimentally perturbed miRNAs in cancer cell lines.
• Applied to colorectal cancer data, the method inferred relevant perturbed miRNA regulatory networks.
• Identified known oncogenic and tumor-suppressive miRNAs, including miR-17, miR-26, and miR-145, in colorectal cancer.

Conclusions:

• The network propagation method effectively leverages network effects for miRNA target gene inference.
• This approach provides a valuable tool for identifying perturbed miRNAs and their key targets in biological processes.
• The method enhances the analysis of gene expression data for understanding miRNA-mediated regulation.