On single and multiple models of protein families for the detection of remote sequence relationships
2006

Comparing Protein Models for Detecting Relationships

Sample size: 1718 publication Evidence: moderate

Author Information

Author(s): Casbon James A, Saqi Mansoor A

Primary Institution: Queen Mary's School of Medicine and Dentistry

Hypothesis

Does a single superfamily model of a large number of diverse sequences perform better at the detection of remote superfamily members than using multiple domain models built for each individual member of the superfamily?

Conclusion

Overall, multiple models perform better in recognition although single structure-based models display better alignment accuracy.

Supporting Evidence

  • Multiple domain models outperform single superfamily models on average.
  • For 12% of superfamilies, superfamily models perform significantly better than domain models.
  • For 10% of superfamilies, domain models show a similar improvement over superfamily models.

Takeaway

This study looked at different ways to group proteins to see which method helps find relationships between them better. It found that using many models is usually better than just one.

Methodology

The study used profile-profile methods to benchmark the performance of single structure-based superfamily models and multiple domain models.

Limitations

The definition of a superfamily may change and the performance can vary significantly between different superfamilies.

Digital Object Identifier (DOI)

10.1186/1471-2105-7-48

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication