DiffGRN: differential gene regulatory network analysis

Area of Science:

• Genomics
• Systems Biology
• Bioinformatics

Background:

• Understanding gene regulation is key to deciphering complex diseases.
• Differential Network Analysis (DiNA) uses gene regulatory networks (GRNs) to study biological processes.
• Current correlation-based GRN methods lack causal and multivariate effect representation.

Purpose of the Study:

• To propose Differential Gene Regulatory Network (DiffGRN) for inferring differential gene regulation between two groups.
• To address limitations of correlation-based methods in DiNA by incorporating causality and multivariate effects.
• To identify differential gene regulators for a better understanding of disease mechanisms.

Main Methods:

• Inferring gene regulatory networks for two groups using Random LASSO.
• Developing a significance test to identify differential gene regulations.
• Comparing DiffGRN with the correlation-based method DINGO using simulations.

Main Results:

• DiffGRN effectively captures multivariate gene effects and identifies causal relationships.
• Simulation experiments demonstrate superior performance of DiffGRN over DINGO.
• Application to asthma gene expression data identified known gene regulations like ADAM12 and RELB.

Conclusions:

• DiffGRN offers an advanced approach for differential gene regulatory network analysis.
• The method enhances the discovery of causal and multivariate gene interactions.
• DiffGRN holds promise for advancing our understanding of gene regulation in diseases like asthma.