Nanopore sequencing enables high-resolution analysis of resistance determinants and mobile elements in the human gut microbiome

Abstract

The analysis of information rich whole-metagenome datasets acquired from complex microbial communities is often restricted by the fragmented nature of assembly from short-read sequencing.The availability of long-reads from third-generation sequencing technologies (e.g. PacBio or Oxford Nanopore) can help improve assembly quality in principle, but high error rates and low throughput have limited their application in metagenomics. In this work, we describe the first hybrid metagenomic assembler which combines the advantages of short and long-read technologies, providing an order of magnitude improvement in contiguity compared to short read assemblies, and high base-pair level accuracy. The proposed approach (OPERA-MS) integrates a novel assembly-based metagenome clustering technique with an exact scaffolding algorithm that can efficiently assemble repeat rich sequences. Based on evaluations with defined in vitro communities and virtual gut microbiomes, we show that it is possible to assemble near complete genomes from metagenomes with as little as 9× long read coverage, thus enabling high quality assembly of lowly abundant species (<1%). Furthermore, OPERA-MS's fine-grained clustering is able to deconvolute and assemble multiple genomes of the same species in a single sample, allowing us to study the complex dynamics of the human microbiome at the sub-species level.Applying nanopore sequencing to gut metagenomes of patients undergoing antibiotic treatment, we show that long reads can be obtained from stool samples in clinical studies to produce more meaningful metagenomic assemblies (up to 200× improvement over short-read assemblies), including the closed assembly of >80 putative plasmid/phage sequences and a 263kbp jumbo phage. Our results highlight that high-quality hybrid assemblies provide an unprecedented view of the gut resistome in these patients, including strain dynamics and identification of novel plasmid sequences.

Cartography of opportunistic pathogens and antibiotic resistance genes in a tertiary hospital environment

“…The use of Illumina shotgun metagenomics by Brooks et al was key to characterizing strain polymorphisms and relatedness of pathogens in low diversity neonatal ICU environments 26 . Several limitations remain for the use of shotgun metagenomics in general, including low microbial biomass, the presence of nosocomial strains at low relative abundances, the presence of multiple strains, computational constraints in strain-level analysis 27 , and the shortcomings of short-read metagenomics for assembling high-contiguity, strainresolved genomes for detailed genetic analysis 24,26,28,29 .…”

mosaicFlye: Resolving long mosaic repeats using long error-prone reads

“…The ongoing large-scale sequencing projects, such as the Earth Biogenome Project (Lewin et al, 2018) and Telomere-2-Telomere project (Miga et al, 2019), use long-read technologies to generate complete genomes. At the same time, large bacterial sequencing projects (such as NCTC 3000) and metagenomic projects (Bertrand et al, 2019) use long reads to dramatically improve bacterial assemblies. However, since two algorithmic problems (assembling long mosaic repeats and long tandem repeats) remain unsolved, these projects have to complement long read generation by rather expensive complementary technologies such as optical maps, synthetic long reads, and Hi-C.…”

Human gut mobileome during antibiotic therapy. A trajectory-based approach to analysis of the metagenomic time-series datasets

“…Therefore the application of long read sequencing technology (e.g. Oxford Nanopore) in future studies, might enable more detailed plasmid analysis and phageome analysis without the need for independent sequencing and purification of VLPs (Bertrand et al, 2018). However, the error rate so far is too high to use the k-mer based plasmid and phage detection methods.…”