BubbleGun

BubbleGun detects and analyzes bubble and superbubble structures in de Bruijn genome graphs to identify polymorphic loci, bubble chains, and structural variation in graph-based genome representations.

Key Features:

• Bubble and superbubble detection: Identifies individual bubbles and superbubbles within de Bruijn graphs that indicate polymorphic regions.
• Bubble chain reporting: Reports runs of linearly connected bubble structures (bubble chains) to capture contiguous variant loci.
• Performance optimization: Implements optimizations for rapid detection of bubbles and superbubbles in large graphs.
• Scalability demonstration: Can process a human-sample de Bruijn graph of approximately 2 million nodes and report all bubbles in under 30 minutes.
• Comparative performance: Exhibits significant speed advantages over vg’s snarl detection on larger datasets.
• Applicability to assembly-derived graphs: Operates on de Bruijn graphs generated from de novo assemblies, including assemblies produced with third-generation sequencing technologies.
• Pangenome relevance: Applies to graph-based pangenomes for capturing population-level genetic diversity and structural variants.

Scientific Applications:

• De novo genome assembly analysis: Detection of polymorphic loci and structural variation in de Bruijn graphs from de novo assemblies of larger genomes.
• Pangenome studies: Identification and characterization of variants within pangenome graph representations to capture population diversity.
• Structural variation discovery: Localization and characterization of structural variants and polymorphic regions represented as bubbles and superbubbles.
• Variant connectivity analysis: Analysis of arrangements and connectivity of variant loci via reported bubble chains.

Methodology:

Rapid detection of bubbles and superbubbles in de Bruijn genome graphs and reporting of runs of linearly connected bubble structures (bubble chains), with performance optimizations demonstrated in comparison to vg’s snarl detection.