ActSeek: fast and accurate search algorithm of active sites in alphafold database

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• Protein function is crucial for advancements in medicine, biotechnology, and material science.
• Current protein mining algorithms focus on sequence or structure, neglecting active site importance.
• Protein activity, especially for enzymes and drug targets, depends on active site residues.

Purpose of the Study:

• To introduce ActSeek, a novel computational tool for mining protein structural databases.
• To enable the discovery of proteins with specific active site environments.
• To leverage computer vision principles for efficient protein structure analysis.

Main Methods:

• ActSeek utilizes a computer vision-inspired approach for fast database searching.
• The program identifies proteins based on similarity to a seed protein's active site.
• ActSeek is applied to mine the AlphaFold database for proteins with desired active site characteristics.

Main Results:

• ActSeek successfully identifies proteins with similar active site environments.
• Demonstrated potential in discovering enzymes for biodegradable plastic production and degradation.
• Identified potential off-target proteins for common drug molecules, highlighting drug safety implications.

Conclusions:

• ActSeek offers a powerful new method for functional protein discovery.
• The tool can accelerate research in areas like biocatalysis and drug development.
• ActSeek facilitates the identification of proteins for environmental and medical applications.