Using indirect protein interactions for the prediction of Gene Ontology functions
2007

Using Indirect Protein Interactions to Predict Gene Functions

Sample size: 7 publication Evidence: high

Author Information

Author(s): Chua Hon Nian, Sung Wing-Kin, Wong Limsoon

Primary Institution: National University of Singapore

Hypothesis

Can indirect protein interactions improve the prediction of Gene Ontology functions across multiple genomes?

Conclusion

FS-Weighted Averaging effectively utilizes indirect interactions to enhance the inference of protein functions from protein interactions.

Supporting Evidence

  • FS-Weighted Averaging consistently outperformed Neighbor Counting and Chi-Square across all three categories of the Gene Ontology.
  • Indirect interactions provided significant additional coverage over annotations that could not be inferred from sequence homology.
  • The method showed robustness against noisy interaction data.

Takeaway

This study shows that looking at how proteins interact with other proteins can help us understand their functions better, especially when we consider indirect connections.

Methodology

The study used FS-Weighted Averaging to analyze protein-protein interactions across seven genomes and compared its performance with Neighbor Counting and Chi-Square methods.

Limitations

The performance of FS-Weighted Averaging is less significant in sparse interaction networks and when interactions are randomly deleted.

Participant Demographics

The study analyzed protein interactions from seven different genomes: Saccharomyces cerevisiae, Drosophila melanogaster, Caenorhabditis elegans, Arabidopsis thaliana, Rattus norvegicus, Mus musculus, and Homo sapiens.

Digital Object Identifier (DOI)

10.1186/1471-2105-8-S4-S8

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication