Fishing with (Proto)Net: A Method for Protein Target Selection
Author Information
Author(s): Michal Linial
Primary Institution: The Hebrew University
Hypothesis
The probability of a cluster without a solved protein belonging to a new superfamily correlates with the distance in the ProtoNet hierarchy.
Conclusion
The study presents a systematic methodology for selecting target proteins likely to adopt new structural superfamilies or folds.
Supporting Evidence
- Over 120,000 proteins are archived in the Swiss-Prot database.
- Only a small fraction (<5%) of newly solved structures have been identified as new folds.
- The FSA-based method resulted in a success rate of about 80% in separating new superfamilies from known ones.
Takeaway
This study helps scientists choose which proteins to study by predicting which ones might have new shapes that we haven't seen before.
Methodology
The study uses a computational-statistical method based on hierarchical clustering of protein sequences to prioritize proteins for structural determination.
Potential Biases
The selection process may favor certain types of proteins over others, potentially leading to an unbalanced representation of protein structures.
Limitations
The method may not account for practical issues such as protein solubility and the need for additional entities for protein function.
Statistical Information
P-Value
80%
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website