InSite: A Method for Predicting Protein-Protein Interaction Binding Sites
Author Information
Author(s): Wang Haidong, Segal Eran, Ben-Hur Asa, Li Qian-Ru, Vidal Marc, Koller Daphne
Primary Institution: Stanford University
Hypothesis
InSite can accurately predict specific binding regions of interacting protein pairs using high-throughput data.
Conclusion
InSite significantly outperforms existing methods in predicting protein-protein interaction binding sites and provides insights into disease mechanisms.
Supporting Evidence
- InSite predicted binding sites were validated against known structures in the Protein Data Bank.
- 90% of the top 50 predicted binding sites were confirmed by experimental data.
- The method integrates multiple sources of evidence, improving prediction accuracy.
Takeaway
InSite is a computer program that helps scientists figure out where two proteins stick together, which is important for understanding diseases.
Methodology
InSite uses a probabilistic model that integrates various data sources to predict binding sites for protein pairs.
Potential Biases
The model may be biased if the training data is not representative of the true interaction landscape.
Limitations
InSite may not identify binding sites not covered by known motifs and cannot distinguish between structural and binding motifs.
Participant Demographics
The study includes data from both Saccharomyces cerevisiae and human proteins.
Statistical Information
P-Value
0.001
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website