Collaborative Filtering Approach for Protein-Protein Docking
Author Information
Author(s): Bourquard Thomas, Bernauer Julie, Azé Jérôme, Poupon Anne
Primary Institution: Bioinformatics Group, INRIA AMIB, Laboratoire de Recherche en Informatique, Université Paris-Sud, Orsay, France
Hypothesis
Can a collaborative filtering approach improve the accuracy of protein-protein docking scoring functions?
Conclusion
The study demonstrates that the collaborative filtering method can effectively rank near-native conformations in protein-protein docking.
Supporting Evidence
- The new scoring function was evaluated using 10-fold cross-validation.
- For 10 out of 12 targets, a near-native conformation was found in the top 10 ranked solutions.
- The method significantly enriched the 100 best-ranking conformations with near-native structures.
Takeaway
This study shows a new way to help scientists find how proteins stick together by using smart computer methods to rank their shapes.
Methodology
The study used a collaborative filtering method combined with genetic algorithms to rank protein conformations based on their predicted interactions.
Limitations
The method struggles with trimers and may not effectively identify multiple interfaces.
Statistical Information
P-Value
0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website