Fig 1.
Illustrated overview of our recommended approach to perfect bacterial whole-genome assembly.
Fig 2.
Examples of pitfalls in bacterial genome assembly and polishing.
(A) A. baumannii J9 [47] contains one large 145-kbp plasmid (blue) and one small 6-kbp plasmid (red). The small plasmid is missing from an ONT-only assembly of this genome (left). However, it assembled completely in an Illumina-only assembly (right), enabling its recovery. (B) IS481 is a repeat in the B. pertussis Tohama I genome [53]. Due to its high copy-number, some errors in this repeat are not fixable using paired-end Illumina reads and short-read polishers. (C) If a genome contains a very long repeat, as is the case with M. smegmatis mc2155 [51], typical ONT read lengths of approximately 20 kbp may not be sufficient for complete assembly. (D) As occurred with Haemophilus M1C132_1 and K. oxytoca MSB1_2C [6], read demultiplexing errors can cause a deeply sequenced replicon in one genome (left) to erroneously appear in the assembly of another genome from the same sequencing run (right). (E) ONT sequencing of K. pneumoniae INF277 [54] contained a near-50:50 mixture of fim switch orientations, causing problems during long-read and short-read polishing. (F) S. aureus JKD6159 [7] read sets contained structural heterogeneity around the ΦSa3 bacteriophage sequence (left), causing an incomplete Flye assembly graph (right).