Structural similarity to link sequence space: new potential superfamilies and implications for structural genomics

Area of Science:

• Structural biology
• Bioinformatics
• Computational biology

Background:

• Protein structure determination now often precedes functional characterization.
• Structure-based homology assignment is crucial for inferring protein function.
• Sequence similarity after structure-based alignment is a strong indicator of homology.

Purpose of the Study:

• To develop a method for predicting distant homologous relationships by merging protein structure and sequence databases.
• To identify novel homologous protein families and superfamilies.
• To infer protein function for proteins with known structures but unknown functions.

Main Methods:

• Utilized the Structural Classification of Proteins (SCOP) database to link sequence alignments from SMART and Pfam.
• Developed new sequence alignments leveraging known three-dimensional protein structures.
• Extended Murzin's method to statistically assess sequence identities after structural alignment.

Main Results:

• Successfully linked several distantly related protein sequence families with confidence.
• Identified new potential protein superfamilies based on conserved structural features and substrate binding sites.
• Demonstrated a robust method for inferring homology and function from structural data.

Conclusions:

• The integrated approach effectively predicts homologous relationships and infers function for proteins with known structures.
• The findings have significant implications for Structural Genomics initiatives and improving sequence comparison algorithms.
• This method enhances our ability to assign function in the post-genomic era.