Section: New Results

Representations of de Bruijn graphs

We designed the first practical data structure for representing large de Bruijn graphs, which supports insertions and deletions of nodes [3]. This work concerns de novo assembly and several other $k$-mer-related bioinformatics problems. The representation of de Bruijn graphs is a transversal bioinformatics question that has enjoyed recent applications in genome, metagenome and transcriptome assembly and quantification. To this date, efficient data structures were essentially static. In this work we provided an implementation of a dynamic data structure that combines perfect hashing, Karp-Rabin hashing, and forests. Practical tests show that this structure is highly competitive with the state of the art.