Understanding Protein-Protein Interaction Networks
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)
Want to read the original?
Access the complete publication on the publisher's website