Predicting Domain-Domain Interactions in Proteins
Author Information
Author(s): Lee Hyunju, Deng Minghua, Sun Fengzhu, Chen Ting
Primary Institution: University of Southern California
Hypothesis
Can integrating multiple biological data sets improve the prediction of domain-domain interactions?
Conclusion
The study demonstrates that using a Bayesian approach to integrate various biological data sets significantly enhances the accuracy of predicting domain interactions.
Supporting Evidence
- The integration of multiple biological data sets significantly increases the coverage and accuracy of predicted domain interactions.
- A total of 2,391 high-confidence domain interactions were identified.
- The accuracy of predicted domain interactions was validated against experimentally obtained interactions.
Takeaway
This study shows how scientists can use different types of biological data to better guess how parts of proteins interact with each other.
Methodology
The study combines protein interaction data from multiple species, molecular sequences, and gene ontology using a Bayesian approach to predict domain-domain interactions.
Limitations
The study does not include all interaction data from all species and the protein interaction data sets used are incomplete and may contain false positives.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website