Recovering Escherichia coli Plasmids in the Absence of Long-Read Sequencing Data

Abstract

The incidence of infections caused by multidrug-resistant E. coli strains has risen in the past years. Antibiotic resistance in E. coli is often mediated by acquisition and maintenance of plasmids. The study of E. coli plasmid epidemiology and genomics often requires long-read sequencing information, but recently a number of tools that allow plasmid prediction from short-read data have been developed. Here, we reviewed 25 available plasmid prediction tools and categorized them into binary plasmid/chromosome classification tools and plasmid reconstruction tools. We benchmarked six tools (MOB-suite, plasmidSPAdes, gplas, FishingForPlasmids, HyAsP and SCAPP) that aim to reliably reconstruct distinct plasmids, with a special focus on plasmids carrying antibiotic resistance genes (ARGs) such as extended-spectrum beta-lactamase genes. We found that two thirds (n = 425, 66.3%) of all plasmids were correctly reconstructed by at least one of the six tools, with a range of 92 (14.58%) to 317 (50.23%) correctly predicted plasmids. However, the majority of plasmids that carried antibiotic resistance genes (n = 85, 57.8%) could not be completely recovered as distinct plasmids by any of the tools. MOB-suite was the only tool that was able to correctly reconstruct the majority of plasmids (n = 317, 50.23%), and performed best at reconstructing large plasmids (n = 166, 46.37%) and ARG-plasmids (n = 41, 27.9%), but predictions frequently contained chromosome contamination (40%). In contrast, plasmidSPAdes reconstructed the highest fraction of plasmids smaller than 18 kbp (n = 168, 61.54%). Large ARG-plasmids, however, were frequently merged with sequences derived from distinct replicons. Available bioinformatic tools can provide valuable insight into E. coli plasmids, but also have important limitations. This work will serve as a guideline for selecting the most appropriate plasmid reconstruction tool for studies focusing on E. coli plasmids in the absence of long-read sequencing data.

Dynamics of extended-spectrum cephalosporin resistance genes in Escherichia coli from Europe and North America

“…A total of 313 typeable ESC-R plasmids were recovered by MOB-suite and RFPlasmid, representing 27.6% of all predicted plasmid contigs associated with ESC-R genes. Although relatively low, this percentage is comparable to that obtained in a recent study 56 ; the remaining plasmid contigs were not typeable as they did not contain replicon sequences.…”

“…Recovery of fully contiguated plasmid sequences from short-read data is very challenging due to the frequent occurrence of repeat elements that impede assembly of complete plasmid or chromosome sequences 56 . Despite this limitation, diverse bioinformatic tools are available for plasmid detection, classification and reconstruction 34,57,58 , with a recent study identifying MOB-suite as the only tool to reconstruct correctly the majority of E. coli plasmids in a test set 56 . Using MOB-suite, we were able to recover typeable ESC-R plasmids, but it over-represented the proportion of ESC-R genes carried on the chromosome.…”

Genomic analysis of the international high-risk clonal lineage Klebsiella pneumoniae sequence type 395

“…The genes of interest cannot be directly assigned to a particular plasmid if found in different contigs than the plasmid marker genes. Assembly of AMR gene-carrying plasmids is especially problematic because they are usually large in size, low in copy number and frequently built as mosaic-like structures, containing mobile components that can be found in different plasmid backbones [ 93 ]. Thus, several software tools have been developed that use graph-based or reference-based approaches or their combination.…”

“…MOB-suite allows “aggregation” of input sequence contigs into putative plasmids according to the reference hits obtained from the reference database of clustered complete plasmid sequences and determination of plasmid types using marker sequence databases of known replicon and relaxase genes. This software has been successfully used to characterize AMR gene-carrying plasmids in many large-scale bacterial population genetics studies, and in a recent study comparing the performance of different available plasmid prediction tools, has been shown to be the only tool to correctly reconstruct the majority of Escherichia coli plasmids, including large ESBL-encoding plasmids [ 93 ]. All available bioinformatics approaches, however, have important limitations and none can correctly reconstruct all plasmids from draft genome assemblies.…”

The effect of sequencing and assembly on the inference of horizontal gene transfer on chromosomal and plasmid phylogenies

“…Third, it would exceed the scope of our study to include all of the (25 or more) different methods that have been developed to detect plasmids from WGS data [ 53 ]. Our method of BLASTing against the PlasmidFinder database to detect putative plasmid sequences will return a lower bound on the plasmid content for a specific strain.…”