The use of Gene Ontology terms for predicting highly-connected 'hub' nodes in protein-protein interaction networks
2008

Predicting Hub Proteins in Protein Interaction Networks

Sample size: 21784 publication Evidence: moderate

Author Information

Author(s): Hsing Michael, Byler Kendall Grant, Cherkasov Artem

Primary Institution: University of British Columbia

Hypothesis

Can Gene Ontology terms be used to predict highly-connected hub proteins in protein-protein interaction networks?

Conclusion

The developed hub classifier accurately identifies highly-connected proteins that share specific functional properties reflected in their Gene Ontology annotations.

Supporting Evidence

  • The hub classifier was validated on external datasets, demonstrating high accuracy in predicting hub proteins.
  • Sensitivity and specificity of the classifier were consistently high across training and testing sets.
  • The classifier outperformed traditional methods for predicting protein interactions.

Takeaway

This study created a tool that helps scientists find important proteins in cells that interact a lot with other proteins, which can help in drug discovery.

Methodology

A hub protein classifier was developed using boosting trees based on Gene Ontology annotations and protein interaction data from multiple species.

Limitations

The performance of the hub classifier relies on the availability of Gene Ontology annotations, which vary among species.

Participant Demographics

Data from Escherichia coli, Saccharomyces cerevisiae, Drosophila melanogaster, and Homo sapiens were used.

Statistical Information

P-Value

p<0.05

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1752-0509-2-80

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication