Practical dynamic de Bruijn graphs.
other
Where this comes from
- Record sourced from PubMed, PMID 29939217.
- Also identified by DOI 10.1093/bioinformatics/bty500.
- No licence information is recorded for this record.
- Because redistribution is not established, this page shows the abstract only. Follow the links below for the full text.
Abstract
The de Bruijn graph is fundamental to the analysis of next generation sequencing data and so, as datasets of DNA reads grow rapidly, it becomes more important to represent de Bruijn graphs compactly while still supporting fast assembly. Previous implementations of compact de Bruijn graphs have not supported node or edge deletion, however, which is important for pruning spurious elements from the graph. Belazzougui et al. (2016b) recently proposed a compact and fully dynamic representation, which supports exact membership queries and insertions and deletions of both nodes and edges. In this paper, we give a practical implementation of their data structure, supporting exact membership queries and fully dynamic edge operations, as well as limited support for dynamic node operations. We demonstrate experimentally that its performance is comparable to that of state-of-the-art implementations based on Bloom filters. Our source-code is publicly available at https://github.com/csirac/dynamicDBG under an open-source license.
Medical subject headings
- Algorithms
- High-Throughput Nucleotide Sequencing
- Sequence Analysis, DNA
- Software