Succinct dynamic de Bruijn graphs.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• The de Bruijn graph is crucial for high-throughput sequencing data analysis.
• Efficient construction and storage are vital for population-scale studies.
• Existing methods lack efficient, dynamic update capabilities, requiring complete reconstruction.

Purpose of the Study:

• To present DynamicBOSS, a novel succinct de Bruijn graph representation.
• To enable efficient and mutable graph construction for large-scale population studies.
• To support unlimited additions and deletions of nodes and edges.

Main Methods:

• Developed DynamicBOSS, a succinct de Bruijn graph data structure.
• Implemented dynamic capabilities for node and edge manipulation.
• Compared DynamicBOSS performance against existing methods on large datasets.

Main Results:

• DynamicBOSS supports unlimited additions and deletions of nodes and edges.
• It is applicable to very large datasets (over 15 billion k-mers).
• Outperforms competing dynamic methods like FDBG and BiFrost in scalability and functionality.

Conclusions:

• DynamicBOSS provides an efficient and mutable solution for de Bruijn graphs in population studies.
• Enables dynamic analysis of large-scale sequencing data.
• Addresses limitations of existing static and less scalable dynamic graph structures.