Fig. 2
From: Graph mining for next generation sequencing: leveraging the assembly graph for biological insights
Focus assembly and analysis pipeline general overview. a Read preprocessor. The 5’ and 3’ read ends are trimmed using quality values and/or by a fixed length specified by the user. Reverse complements of reads are generated and the processed read data set is split into subsets for processing by the parallel read aligner. b Parallel pairwise read alignment. The reads subsets are pairwise aligned using a suffix-array seed based search and extend method. c Multilevel graph set. Iterative heavy edge matching and node merging is used to create a set of graphs. d Hybrid graph. Best representative nodes are selected at each graph level using partial assembly to create the hybrid graph G0. e Hybrid graph trimming. Transitive edges and redundant nodes are trimmed from G’ 0. Figure 2 (a-e) are a simplified overview, please see Figs. 3 and 4 and corresponding methods section for more details regarding the construction of the multilevel graph set, hybrid graph, and trimming of the hybrid graph