BLANT

BLANT performs local alignment of biological networks by sampling small induced k-node graphlets and using a seed-and-extend heuristic to index and compare network substructures.

Key Features:

• Graphlet Sampling: Samples small, induced k-node graphlets from input networks to represent local topology.
• High-throughput Sampling: Generates millions of graphlet samples within seconds for large-scale network analysis.
• Seed-and-Extend Indexing: Constructs an index of graphlet-derived seeds that serves as a foundation for seed-and-extend local alignments.
• Versatile Outputs: Produces distributions of graphlets or their orbits, graphlet degree vectors and graphlet orbit degree vectors, and an index usable for local alignment workflows.
• ORCA Compatibility: Outputs include graphlet degree vectors and graphlet orbit degree vectors compatible with ORCA.

Scientific Applications:

• Network classification: Classifies networks based on structural features derived from graphlet distributions and orbit statistics.
• Network quantification and alignment: Quantifies network structure and supports both local and global alignment of biological networks.
• Topology–function analysis: Identifies relationships between network topology and biological function.
• Sequence-free phylogeny: Constructs taxonomic trees without relying on sequence data using structural network signatures.
• Functional similarity of PPI networks: Detects functional similarity between protein–protein interaction networks, exemplified by yeast and human PPI comparisons using the indexing mode.

Methodology:

BLANT samples induced k-node graphlets from input networks and applies a BLAST-like seed-and-extend heuristic that identifies graphlet-derived seeds and extends them into longer local alignments, using the sampled graphlets to build an index.