Inferring Gene Network Rewiring by Combining Gene Expression and Gene Mutation Data.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Gene dependency network rewiring is key to understanding disease progression.
• Existing differential network analysis methods are often limited to specific data types.
• High-throughput technologies generate diverse gene activity measurements (e.g., mRNA expression, DNA mutation).

Purpose of the Study:

• To develop a novel computational model for inferring differential gene networks by integrating multiple data types.
• To explore similarities and differences between networks derived from distinct genomic data.
• To identify gene network rewiring associated with ovarian cancer platinum resistance.

Main Methods:

• Developed a new differential network inference model combining gene expression and DNA mutation data.
• Employed a group bridge penalty function to learn similarities and differences across data types.
• Validated the method using simulation studies and applied it to ovarian cancer data.

Main Results:

• The proposed method outperforms existing approaches in simulation studies.
• Identified both common and unique differential network edges between gene expression and mutation data.
• Discovered hub genes in differential networks critical for ovarian cancer drug resistance.

Conclusions:

• The integrated approach provides a more comprehensive understanding of gene network rewiring.
• The model effectively handles multi-modal genomic data for differential network analysis.
• Key genes identified play significant roles in ovarian cancer platinum resistance, offering potential therapeutic targets.