Pushing the limits of de novo genome assembly for complex prokaryotic genomes harboring very long, near identical repeats

Abstract

Generating a complete, de novo genome assembly for prokaryotes is often considered a solved problem. However, we here show that Pseudomonas koreensis P19E3 harbors multiple, near identical repeat pairs up to 70 kilobase pairs in length, which contained several genes that may confer fitness advantages to the strain. Its complex genome, which also included a variable shufflon region, could not be de novo assembled with long reads produced by Pacific Biosciences’ technology, but required very long reads from Oxford Nanopore Technologies. Importantly, a repeat analysis, whose results we release for over 9600 prokaryotes, indicated that very complex bacterial genomes represent a general phenomenon beyond Pseudomonas. Roughly 10% of 9331 complete bacterial and a handful of 293 complete archaeal genomes represented this ‘dark matter’ for de novo genome assembly of prokaryotes. Several of these ‘dark matter’ genome assemblies contained repeats far beyond the resolution of the sequencing technology employed and likely contain errors, other genomes were closed employing labor-intense steps like cosmid libraries, primer walking or optical mapping. Using very long sequencing reads in combination with assembly algorithms capable of resolving long, near identical repeats will bring most prokaryotic genomes within reach of fast and complete de novo genome assembly.

A megaplasmid family responsible for dissemination of multidrug resistance inPseudomonas

“…The association of key adaptive traits with duplicated regions has also been reported for the p1 megaplasmid-carrying Pseudomonas koreensis strain P19E3 (64) , with genes encoding heavy metal resistance, aromatic compound degradation, and DNA repair occurring within large repeats in the genome. Consistent with our observations, repetitive genes in p1 were identified close to transposase and phage genes suggesting a key role for mobile elements like transposons in shaping these large genome duplications (64) .…”

“…The pRB16 megaplasmid was identified in P. citronellolis SJTE-3, an estrogen and polycyclic aromatic hydrocarbon degrading bacterium isolated from active sludge at a wastewater treatment plant in China (75) . P. koreensis P19E3, harbouring the megaplasmid p1, was isolated from healthy marjoram ( Origanum marjorana ) leaf material during an isolation survey on an organic herb farm (Boppelsen, Switzerland) in 2014 (64) . The presence on p1 of a cluster of genes encoding copper resistance suggests that whilst members of the megaplasmid family share a common backbone, they are flexible in the adaptive traits that they carry.…”

“…Rarefaction and accumulation curves, used to estimate the core and pan-genome size of the megaplasmid family, were indicative of an open pan-genome with a well-defined core, and suggested that the addition of more genomes was unlikely to impact much further on the size of the core genome ( Figure 3). Further analysis was carried out to distinguish between the strict core genome (present in all 15 megaplasmids), and other pan-genome compartments (40) (Figure 4, Supplementary Figure S5), indicating that the greatest number of gene clusters corresponded to unique, or nearly unique genes, but with numbers varying between the different megaplasmids from only one to 93 unique genes (Supplementary Figure S5, Supplementary Table 4), the latter being found in plasmid p1, a plasmid present in a P. koreensis isolate from healthy marjoram leaf material (64) . Genes unique to plasmid p1 include a large cluster related to copper resistance and catabolic genes for the conversion of 4-hydroxy benzoate ( pobA , phbH ), or vanillate ( vanA , vanB , iacF , vanK ) to protocatechuate, which can then be metabolised to central metabolism intermediates via the β-ketoadipate pathway.…”

A proteogenomic resource enabling integrated analysis ofListeriagenotype-proteotype-phenotype relationships

“…An analysis of the repeat complexity of all publicly available, completely sequenced genomes of L. monocytogenes strains (status: March, 2018) revealed that almost 95% of the roughly 150 strains are so-called "class I" genomes, which are straightforward to assemble (few repeats, none longer than the rDNA operons of up to 7 kb) [17]. In contrast, eight strains also had few repeats overall, but those present were up to 11 kb in length [17].…”

“…An analysis of the repeat complexity of all publicly available, completely sequenced genomes of L. monocytogenes strains (status: March, 2018) revealed that almost 95% of the roughly 150 strains are so-called "class I" genomes, which are straightforward to assemble (few repeats, none longer than the rDNA operons of up to 7 kb) [17]. In contrast, eight strains also had few repeats overall, but those present were up to 11 kb in length [17]. Using long-read Pacific Biosciences (PacBio) sequencing data (including a BluePippin size selection step; see Materials and Methods) and the assembly algorithm HGAP3 [50], we were able to de novo assemble one complete chromosome for EGD-e (2.94 Mbp) and one for ScottA (3.03 Mbp) with PacBio read coverages of 260x and 280x, respectively (Fig 2).…”