GrAfSS

GrAfSS performs graph-theoretical substructure similarity searching and comparison of protein and RNA three-dimensional structures to identify and compare amino acid side-chain and base-arrangement motifs.

Key Features:

• Graph-theoretical algorithms: Integrates five algorithms—ASSAM, SPRITE, IMAAAGINE, NASSAM, and COGNAC—for graph-based substructure searching.
• Substructure search capabilities: Searches for specific three-dimensional arrangements, including ligand binding sites and other 3D motifs, across structures in the Protein Data Bank (PDB).
• Molecular focus: Targets amino acid side chains in proteins and base arrangements in RNA for detailed substructural analysis.
• Functional annotation: Enables functional annotation of experimentally determined and computationally predicted structures, including AlphaFold2 models, by comparison to known motifs.
• Visualization tools: Produces computed substructural superpositions viewable with NGL viewers for structural interpretation.

Scientific Applications:

• Structural motif identification: Identify conserved three-dimensional substructures to infer functional mechanisms.
• Protein function and ligand interactions: Analyze protein functional sites and ligand-binding arrangements across PDB structures.
• RNA structural analysis: Characterize RNA base-arrangement motifs and study RNA structural dynamics.
• Annotation and validation of predicted models: Annotate and support validation of computationally generated structures such as AlphaFold2 predictions.
• Evolutionary inference: Infer evolutionary relationships through comparison of conserved substructures.

Methodology:

Structures are mapped to graph representations with nodes representing atoms or residues and edges denoting spatial proximity or chemical bonds, and the integrated algorithms perform substructure searches by comparing these graphs against a database of known motifs in the PDB.