Fig. 1

Overview of the simplitig approach. a Textual representations of k-mer sets ordered by the degree of compaction: individual k-mers, maximal unitigs, and maximal simplitigs. Every component of a simplitig subgraph (black arrows) of the de Bruijn graph (all arrows) corresponds to a path, and its spelling constitutes a simplitig (the “Methods” section). b Scheme of all possible simplitig representations according to the degree of compaction. While unitigs (dark gray area) correspond to compaction along non-branching nodes in the associated de Bruijn graph, simplitigs (gray area) can also contain branching nodes. Every step of compaction decreases the number of sequences (NS) and their cumulative length (CL) by 1 and by k − 1, respectively. Maximal simplitigs may not be determined uniquely; the simplitig representation can have multiple local optima, depending on which edges were selected at the branching nodes. c The workflow of simplitigs. Simplitigs represent de Bruijn graphs and carry implicitly the same information as unitigs. de Bruijn graphs are usually computed from either assemblies or weighted de Bruijn graphs. Weighted de Bruijn graphs are typically obtained by k-mer counting and allow removing noise, e.g., low-frequency k-mers, which frequently originate in sequencing errors