Topological ranks reveal functional knowledge encoded in biological networks: a comparative analysis

Area of Science:

• Bioinformatics
• Systems Biology
• Network Science

Background:

• Biological network topology provides crucial insights into biological functions, disease mechanisms, and drug discovery.
• Topological measures are used to infer the biological relevance of network components based on their network positions.
• A critical examination of topology's role in extracting novel biological knowledge has been lacking.

Purpose of the Study:

• To comparatively analyze nine topological measures for biological networks.
• To assess the ability of these measures to rank network components according to implicit functional knowledge.
• To evaluate the performance across different organisms and validation criteria.

Main Methods:

• Comparative analysis of nine topological measures using compact views from induced ranks.
• Validation using both internal criteria and external gold standards.
• Application to six biological networks from yeast, worm, and human.

Main Results:

• A subset of top-performing topological measures was identified for each organism.
• The performance of these measures was consistent across different validation approaches.
• The findings highlight organism-specific topological characteristics relevant for biological knowledge extraction.

Conclusions:

• Topology-based measures can effectively rank biological network components.
• Specific topological measures demonstrate superior performance for knowledge extraction.
• The study provides a framework for selecting appropriate topological measures in systems biology research.