Construction of protein interaction network involved in lung adenocarcinomas using a novel algorithm

Area of Science:

• Bioinformatics
• Systems Biology
• Genomics

Background:

• Investigating disease mechanisms requires more than just differentially-expressed (DE) genes.
• Understanding protein interactions is crucial for describing cellular functions.
• Existing protein interaction network studies yield inconsistent results due to varied methodologies.

Purpose of the Study:

• To develop and evaluate a novel combined algorithm for constructing gene interaction networks.
• To compare the topological properties of networks generated by different methods.
• To identify reliable gene interactions for understanding lung adenocarcinoma.

Main Methods:

• Identified DE genes using the RankProd package in lung adenocarcinoma samples.
• Constructed individual networks using empirical Bayesian (EB) meta-analysis, STRING, WGCNA, and DCGL.
• Developed a novel rank-based algorithm to create a combined gene interaction network.

Main Results:

• Screened 941 DE genes and analyzed topological features of five networks.
• The WGCNA network exhibited small-world properties; the combined network was scale-free.
• Gene pairs from the combined method were enriched in cell cycle and p53 signaling pathways.

Conclusions:

• A combined network approach offers a more credible analysis of gene interactions compared to single methods.
• The novel algorithm provides a new perspective for network-based disease mechanism studies.
• This integrated method enhances the reliability of identifying key gene interactions.