CENTRA: knowledge-based gene contextuality graphs reveal functional master regulators by centrality and fractality

Area of Science:

• Bioinformatics
• Systems Biology
• Computational Biology

Background:

• Traditional gene set enrichment analyses lack single-gene resolution and context flexibility.
• Static gene sets limit the ability to associate genes with specific biological contexts.

Purpose of the Study:

• Introduce CENTRA (Centrality-Based Exploration of Network Topologies from Regulatory Assemblies) for context-aware gene function deciphering.
• Develop a framework to model gene contextuality using topic-specific gene co-occurrence networks.

Main Methods:

• Applied latent Dirichlet allocation to 12,045 abstracts linked to Molecular Signatures Database C2 gene sets.
• Uncovered 27 biological topics and constructed topic-specific networks.
• Computed graph-topological metrics (centrality, fractality, perturbation sensitivity) for genes within networks.

Main Results:

• Topological profiles distinguished known regulators and identified potential functional candidates.
• Revealed context-specific gene roles and prioritized understudied genes based on network signature robustness.
• Developed an interactive browser for dynamic network and functional annotation exploration.

Conclusions:

• CENTRA offers an interpretable and scalable framework for investigating context-dependent gene function.
• The approach provides a novel entry point for hypothesis generation beyond traditional enrichment methods.
• Facilitates dynamic navigation of biological networks and functional annotations.