Cutoff lensing: predicting catalytic sites in enzymes

Area of Science:

• Molecular Biomedicine
• Structural Biology
• Enzymology

Background:

• Predicting function-related amino acids and allosteric binding sites in proteins is vital for molecular biomedicine.
• Identifying catalytic sites in enzymes is essential for understanding protein function and drug targeting.

Purpose of the Study:

• Introduce a simple, light, and computationally inexpensive structure-based method for identifying enzyme catalytic sites.
• Validate the method's efficacy using a large database of annotated enzymes.
• Explore the relationship between protein size and predictive power.

Main Methods:

• Develop a method termed 'cutoff lensing' involving increasing cutoff values in elastic network models.
• Utilize three distinct structure-based indicators to identify catalytic sites.
• Validate the method against a comprehensive database of annotated enzymes.

Main Results:

• Optimal cutoff values were identified, maximizing the recovery of known catalytic sites.
• The predictive power of the cutoff lensing method increases with larger protein structures.
• The method demonstrated effectiveness in recovering known catalytic sites across a diverse enzyme database.

Conclusions:

• Cutoff lensing offers a computationally efficient approach to predict enzyme catalytic sites.
• The method can serve as a valuable complement to existing sequence- and structure-based screening techniques.
• Further development could involve combining indicators for a unified figure of merit and sequential analysis.