Automatic generation of 3D motifs for classification of protein binding sites
2007

Automatic Generation of 3D Motifs for Protein Binding Sites

Sample size: 18 publication Evidence: moderate

Author Information

Author(s): Jean-Christophe Nebel, Pawel Herzyk, David R. Gilbert

Primary Institution: Kingston University and University of Glasgow

Hypothesis

Can we automatically generate 3D motifs of protein binding sites based on consensus atom positions?

Conclusion

The study introduces a new method for generating 3D patterns from protein binding sites that are biochemically meaningful and useful for classification and annotation.

Supporting Evidence

  • The method matched 31% of proteins with adenine-based ligands.
  • 95.5% of the matched proteins were classified correctly.
  • The study revealed complex evolutionary links between ligases and transferases.

Takeaway

The researchers created a way to automatically find patterns in how proteins bind to molecules, which helps us understand their functions better.

Methodology

The method involves generating ligand-specific training sets, comparing binding sites, clustering proteins, and generating consensus 3D patterns.

Potential Biases

Potential bias in clustering due to the selection of training sets.

Limitations

The method may not capture all functional similarities due to the complexity of protein structures.

Digital Object Identifier (DOI)

10.1186/1471-2105-8-321

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication