Figure 1.

The four bacterial assembly strategies examined here. A. Short read only assembly. Short next generation sequencing reads generated by the Illumina or Life Technologies platforms (in blue) are assembled with any number of assembly software packages into contiguous sequences (contigs; in red). Here, the software packages Velvet, Ray, and MIRA were evaluated. B. Hybrid Assembly. In this method of assembly, short, higher accuracy reads from the Illumina/Life Technologies platforms (blue), error correct the long, lower accuracy PacBio reads (in green with red Xs denoting basecalling errors) via alignment. The software package pacBioToCA was used for this purpose. These error-corrected reads are then assembled into contigs (in red) using a software package suited to long read assembly such as the Celera assembler. Assembled contigs can be further scaffolded and gap-filled as in D. C. In PacBio-only assembly, long reads are aligned to each other, enabling self-correction. These self-corrected reads are assembled into contigs and can be further scaffolded and gap-filled as in D. D. Hybrid scaffolding, in this method, preassembled contigs (red), derived from any of the methods laid out in A-C, are scaffolded with long PacBio reads (in green). These long scaffolds, which can contain long strings of “N”s (dotted green lines) in between the contigs, are then gap-filled with the program PBJelly.