SMID: A Database for Small Molecule Binding Site Annotation
Author Information
Author(s): Snyder Kevin A, Feldman Howard J, Dumontier Michel, Salama John J, Hogue Christopher WV
Primary Institution: The Blueprint Initiative
Hypothesis
Can a domain-based approach improve the accuracy of predicting small molecule binding sites on proteins?
Conclusion
SMID-BLAST accurately predicts small molecule targets and binding site residues for proteins, enhancing drug discovery efforts.
Supporting Evidence
- SMID-BLAST predictions were validated against 793 experimental small molecule interactions.
- 472 of the predicted interactions matched the experimental data, indicating a 60% accuracy.
- 344 predictions had over 80% of the binding site residues correctly identified.
Takeaway
This study created a tool that helps scientists find where small molecules can attach to proteins, which is important for drug development.
Methodology
The study used a set of co-crystallized protein-small molecule structures and the RPS-BLAST algorithm to generate predictions.
Potential Biases
Potential for false positives due to transitive annotation errors.
Limitations
The predictions are limited to small molecules that exist in the Protein Data Bank.
Statistical Information
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website