LIGSITEcsc: A Tool for Predicting Ligand Binding Sites
Author Information
Author(s): Huang Bingding, Schroeder Michael
Primary Institution: Technical University Dresden, Germany
Hypothesis
Can the use of the Connolly surface and conservation improve the prediction of ligand binding sites on proteins?
Conclusion
LIGSITEcsc improves the prediction of ligand binding sites by using the Connolly surface and conservation ranking.
Supporting Evidence
- LIGSITEcsc achieved a success rate of 71% for top 1 predictions and 75% for top 3 predictions.
- The method was tested on 210 bound structures and 48 unbound/bound structures.
- LIGSITEcsc outperformed other methods like LIGSITE, PASS, SURFNET, and CAST.
Takeaway
This study created a tool called LIGSITEcsc that helps find where drugs can attach to proteins by looking at their surfaces and how similar those surfaces are across different proteins.
Methodology
The study compared LIGSITEcsc with other algorithms on datasets of protein structures to evaluate its performance in predicting binding sites.
Potential Biases
The study relies on the accuracy of the conservation scores and the datasets used for validation.
Limitations
The method may misclassify some non-binding sites as potential binding sites.
Statistical Information
P-Value
p<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