Predicting Protein Interactions
Author Information
Author(s): Benjamin A. Shoemaker, Anna R. Panchenko
Primary Institution: Computational Biology Branch of the National Center for Biotechnology Information
Hypothesis
Can computational methods effectively predict protein and domain interaction partners?
Conclusion
Computational methods can complement experimental approaches to predict protein interactions, despite limitations in data completeness.
Supporting Evidence
- High-throughput experimental methods produce large amounts of data that are often incomplete.
- Computational methods can help validate experimental data and suggest potential protein interactions.
- Different computational methods have varying degrees of success in predicting interactions.
Takeaway
Scientists can use computer programs to guess how proteins work together, even when they don't have all the information.
Methodology
The review discusses various computational methods for predicting protein interactions, including genomic inference, phylogenetic profiles, and classification methods.
Potential Biases
The reliance on existing experimental data may introduce biases based on the types of proteins studied.
Limitations
The methods rely on incomplete experimental data and assume independent interactions between domains.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website