Related Experiment Videos
Bridging structural biology and genomics: assessing protein interaction data with known complexes
Aled M Edwards1, Bart Kus, Ronald Jansen
1Banting and Best Department of Medical Research, University of Toronto, C.H. Best Institute, 112 College St, Toronto, Ontario, Canada M5G 1L6.
Trends in Genetics : TIG
|September 28, 2002
Summary
Genome-wide protein-protein interaction datasets contain many errors, inconsistent with known 3D structures. Integrating multiple noisy datasets using Bayesian methods significantly reduces the error rate for accurate protein interaction mapping.
Area of Science:
- Molecular Biology
- Structural Biology
- Bioinformatics
Background:
- Genome-wide studies aim to map protein-protein interactions (PPIs).
- The reliability and coverage of experimental methods are crucial for the utility of PPI networks.
- Known macromolecular complexes serve as objective benchmarks for validating interaction data.
Purpose of the Study:
- To assess the consistency of genome-wide PPI datasets with known 3D structures.
- To evaluate inconsistencies within and between PPI datasets and larger complex databases.
- To develop methods for improving the accuracy of PPI data.
Main Methods:
- Comparison of PPIs from genome-wide datasets against 3D structures of RNA polymerase II, Arp2/3, and the proteasome.
- Analysis of inconsistencies among multiple genome-wide PPI datasets.
- Integration of individual PPI datasets using Bayesian approaches.
Main Results:
- A significant fraction of PPIs in genome-wide datasets and literature are inconsistent with known 3D structures.
- Marked inconsistencies were observed when comparing different genome-wide datasets and with a larger set of 174 complexes.
- Bayesian integration of noisy PPI datasets substantially decreased the error rate.
Conclusions:
- Current genome-wide PPI data exhibit considerable inaccuracies when validated against structural information.
- Integrating multiple, individually noisy PPI datasets is essential for reliable interaction mapping.
- Bayesian methods offer a robust approach to enhance the accuracy of PPI networks.