Identifying term relations cross different gene ontology categories.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• The Gene Ontology (GO) organizes gene and gene product functions into three domains: molecular function, biological process, and cellular component.
• Understanding relationships between GO terms across these domains is crucial for advanced biological reasoning.
• Existing methods for calculating cross-domain GO term similarity have limitations, either using only GO data or incomplete gene co-function network information.

Purpose of the Study:

• To propose a novel method, C r o G O2, for measuring similarities between Gene Ontology (GO) terms across different categories.
• To enhance biological reasoning by integrating diverse data sources for improved term similarity calculations.

Main Methods:

• An iterative ranking-based approach (C r o G O2) was developed.
• The method incorporates GO term level information.
• It utilizes both direct and indirect interactions from gene co-function networks.

Main Results:

• C r o G O2 demonstrates superior performance compared to existing methods in measuring cross-category GO term similarities.
• A high-confidence genome-specific term association network was constructed for yeast.
• The identified term linkages are supported by existing literature.

Conclusions:

• The C r o G O2 method provides a more effective way to assess relationships between GO terms across different functional categories.
• The generated yeast term association network offers valuable insights and can be validated through literature.
• The network aids in identifying gene set-related terms, showing overlap with GO enrichment analysis results.