Inferring topology from clustering coefficients in protein-protein interaction networks
2006

Understanding Protein-Protein Interaction Networks

publication Evidence: moderate

Author Information

Author(s): Friedel Caroline C, Zimmer Ralf

Primary Institution: Ludwig-Maximilians-Universität München

Hypothesis

The scale-free nature of protein-protein interaction networks may not accurately reflect the topology of the complete interactome due to sampling errors.

Conclusion

The study suggests that while the correct topology of the interactome cannot be definitively inferred, many topologies can be excluded with high confidence.

Supporting Evidence

  • Sampling with limited bait and edge coverage lowers clustering coefficients significantly.
  • Clustering coefficients observed in protein-protein interaction maps provide a lower bound on the clustering coefficients of complete interactomes.
  • False positive interactions can increase clustering coefficients in randomly clustered networks but decrease them in highly clustered networks.

Takeaway

Scientists studied how missing data in protein interactions can make it hard to understand how proteins connect in a cell, but they found ways to rule out some incorrect ideas about these connections.

Methodology

The study used analytical and simulation methods to investigate the effects of limited sampling on clustering coefficients in protein-protein interaction networks.

Potential Biases

The model assumes that false positive interactions are influenced by the degree of the nodes, which may introduce bias.

Limitations

The study primarily focuses on the effects of sampling and does not account for all possible errors in protein interaction data.

Digital Object Identifier (DOI)

10.1186/1471-2105-7-519

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication