A mixture of feature experts approach for protein-protein interaction prediction
2007

Predicting Protein-Protein Interactions Using Mixture-of-Feature-Experts

Sample size: 30000 publication 10 minutes Evidence: high

Author Information

Author(s): Qi Yanjun, Klein-Seetharaman Judith, Bar-Joseph Ziv

Primary Institution: Carnegie Mellon University

Hypothesis

Can a Mixture-of-Feature-Experts method improve the prediction of protein-protein interactions by utilizing different feature sets for different protein pairs?

Conclusion

The Mixture-of-Feature-Experts method improved upon previous methods for predicting protein-protein interactions in yeast and human cells.

Supporting Evidence

  • The Mixture-of-Feature-Experts method improved prediction accuracy compared to traditional classifiers.
  • The method allows for the evaluation of the contribution of different features to the predictions.
  • The study demonstrated the method's effectiveness using both yeast and human protein interaction data.

Takeaway

Scientists created a smart way to guess which proteins work together by using different types of information for different pairs of proteins.

Methodology

The study used a Mixture-of-Feature-Experts method that combines scores from different classifiers based on the available features for each protein pair.

Potential Biases

Potential bias may arise from the reliance on high-throughput data sources that may not always be reliable.

Limitations

The method relies on the availability and quality of biological datasets, which can be noisy and contain missing values.

Participant Demographics

The study focused on protein interactions in yeast and human cells.

Statistical Information

P-Value

0.9463

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1471-2105-8-S10-S6

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication