Arete - candidate gene prioritization using biological network topology with additional evidence types

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Candidate gene list refinement is crucial for experimental validation after high-throughput analysis.
• Biological data complexity necessitates flexible solutions for comprehensive data utilization.
• Identifying causal genes requires effective integration of diverse data sources.

Purpose of the Study:

• To develop an integrated framework for gene prioritization.
• To link existing network-based methods with novel integrative ranking.
• To enhance the selection of promising candidate genes for experimental verification.

Main Methods:

• Developed a framework integrating two network-based gene prioritization approaches (DIAMOnD, random walk with restart).
• Incorporated an isolation forest-based integrative ranking method.
• Implemented as a Cytoscape app extension for synergy with other analyses.

Main Results:

• Created an accessible framework combining network topology and additional evidence.
• The Cytoscape extension allows simultaneous consideration of multiple data types.
• Demonstrated utility through two disease gene prioritization case studies.

Conclusions:

• Provided efficient reference implementations of DIAMOnD and random walk with restart for Cytoscape.
• Developed an extension to combine algorithm outputs with additional data.
• Showcased the tool's capability in evaluating different gene prioritization approaches for disease research.