3D-blast: 3D protein structure alignment, comparison, and classification using spherical polar Fourier correlations

Area of Science:

• Structural bioinformatics
• Computational biology
• Protein structure analysis

Background:

• Protein structure comparison is crucial for understanding function and evolution.
• Existing methods often rely on sequence alignment, limiting structural comparisons.
• A robust, sequence-independent method is needed for large-scale structural analysis.

Purpose of the Study:

• To develop and validate a novel sequence-independent method for protein structure alignment.
• To enable efficient and accurate 3D structural database searching.
• To assess the performance of the method against established classifications and benchmarks.

Main Methods:

• Utilized three-dimensional spherical polar Fourier (SPF) representations for protein shape.
• Applied SPF to cluster protein structures within the CATH database.
• Performed large-scale database searches using SPF representations and a rotationally-invariant pre-filter.

Main Results:

• The automatic SPF clustering method showed high agreement with expert-curated CATH classifications.
• Database searches demonstrated excellent precision and recall, validated by ROC plot analyses.
• Incorporating a pre-filter significantly reduced query times with minimal accuracy loss.

Conclusions:

• The SPF-based method offers a powerful and efficient tool for sequence-independent protein structure comparison.
• This approach facilitates rapid online 3D structural searches at interactive timescales.
• The method holds significant potential for advancing structural bioinformatics and drug discovery.