Integrating categorical and structural proximity in Disease Ontologies

Area of Science:

• Bioinformatics
• Systems Biology
• Computational Biology

Background:

• Human-curated disease ontologies are vital for clinical decision support but often reflect a reductionist view.
• Understanding disease similarities is crucial for refining classifications and identifying new therapeutic targets.

Purpose of the Study:

• To investigate the relationship between categorical disease proximity in ontologies and structural proximity of disease modules (DMs) in the interactome.
• To develop and validate a network-based methodology for analyzing disease similarities and refining classification systems.

Main Methods:

• Proposing algorithms to automatically induce hierarchical structures from DM proximity relations.
• Comparing the induced hierarchical structures with existing human-curated disease taxonomies.
• Analyzing commonalities and differences between structural and categorical disease similarity.

Main Results:

• Demonstrated a network-based methodology to analyze disease similarities.
• Showcased the ability to refine human disease classification systems.
• Identified promising network areas for discovering new disease-gene interactions.

Conclusions:

• The proposed methodology offers a novel approach to understanding disease relationships beyond traditional classifications.
• This network-based analysis can enhance clinical decision support and guide future research in disease-gene interactions.