An accurate, sensitive, and scalable method to identify functional sites in protein structures

Area of Science:

• Biochemistry
• Computational Biology
• Structural Biology

Background:

• Functional sites dictate protein activity and interactions, serving as primary drug targets.
• Rapid growth in protein sequence and structure data outpaces experimental identification of functional sites.
• Existing methods lack quantitative validation for identifying functional sites computationally.

Purpose of the Study:

• To computationally identify and statistically validate functional sites in protein sequences.
• To assess the accuracy of the Evolutionary Trace (ET) method in pinpointing functional amino acids.
• To provide a quantitative basis for focusing experimental and computational studies on critical protein regions.

Main Methods:

• Development and application of a computational Evolutionary Trace (ET) method.
• Analysis of 86 diverse protein structures, including those from structural genomics.
• Statistical evaluation of the overlap between evolutionarily ranked residues and known functional sites.

Main Results:

• The ET method ranks amino acids by evolutionary importance, revealing statistically significant clustering of top-ranked residues.
• This clustering significantly overlaps with known functional sites across diverse protein structures.
• An automated ET approach accurately identified 70-90% of functional sites based on statistical measures.

Conclusions:

• The overlap between evolutionary trace and functional sites is a recurrent and statistically significant feature.
• The ET method provides a quantitative tool to accurately identify functional sites from sequence and structure data.
• This computational approach can guide structure-function studies, drug design, protein engineering, and functional annotation efforts.