Automated strain separation in low-complexity metagenomes using long reads

Abstract

High-throughput short-read metagenomics has enabled large-scale species-level analysis and functional characterization of microbial communities. Microbiomes often contain multiple strains of the same species, and different strains have been shown to have important differences in their functional roles. Despite this, strain-level resolution from metagenomic sequencing remains challenging. Recent advances on long-read based methods enabled accurate assembly of bacterial genomes from complex microbiomes and an as-yet-unrealized opportunity to resolve strains. Here we present Strainberry, a metagenome assembly method that performs strain separation in single-sample low-complexity metagenomes and that relies uniquely on long-read data. We benchmarked Strainberry on mock communities and showed it consistently produces strain-resolved assemblies with near-complete reference coverage and 99.9% base accuracy. We also applied Strainberry on real datasets for which it improved assemblies generating 27-89% additional genomic material than conventional metagenome assemblies on individual strain genomes. Our results hence demonstrate that strain separation is possible in low-complexity microbiomes using a single regular long read dataset. We show that Strainberry is also able to refine microbial diversity in a complex microbiome, with complete separation of strain genomes. We anticipate this work to be a starting point for further methodological improvements aiming to provide better strain-resolved metagenome assemblies in environments of higher complexities.

Generation of lineage-resolved complete metagenome-assembled genomes by precision phasing

“…Even though bins are often generalized to represent distinct microbial taxonomic units in a sample, they are rarely assumed to accurately represent true, genetically distinct microbial populations in a sample. This problem has been addressed by multiple studies 11,12 , and precise definitions for individual, highly resolved MAGs remain contextual to each study. Similar to one of these studies 11 , we focus on generating separate representative reference genomes for distinct microbial lineages within an individual metagenome, which we define as "lineage-resolved MAGs" 1,13 .…”

“…Furthermore, these workflows are designed primarily to identify strain lineages from alignments of short-read data and do not capture variant linkage data from longer read datasets. A recent attempt to adapt uncorrected long reads to this purpose requires the use of manual curation and a priori estimates of strain numbers in order to achieve optimal results 12 . An intuitive and automated method to generate lineage-resolved complete MAGs is needed for analysis of more complex metagenome communities in order to reduce the time required to validate results.…”

Strainline: full-length de novo viral haplotype reconstruction from noisy long reads

“…We recall that Strainline is unique insofar as it is the first approach to determine the haplotype/strain-specific genomes of viruses from long reads de novo. For the sake of a meaningful comparison, we chose long read de novo assemblers that are designed to deal with mixed samples (in other words, designed for metagenome assembly), such as Canu Koren et al (2017) and metaFlye Vicedomini et al (2021), on the one hand, and generic (consensus) de novo assemblers, such as wtdbg2 Ruan and Li (2020) and Shasta Shafin et al (2020) on the other hand. Of those, we subsequently excluded metaFlye, because it failed to perform the assemblies.…”

“…In addition, metaFlye, originally designed to perform assembly of metagenomes, operates at the level of species Vicedomini et al (2021), so neglects to resolve individual genomes at the level of strains.…”

Recovery of high quality metagenome–assembled genomes from full–scale activated sludge microbial communities in a tropical climate using longitudinal metagenome sampling

“…Despite impressive accomplishments, the MAG approach still harbours many challenges and limitations. By nature, short read metagenome assemblies remain highly fractionated, resulting from the limited ability of short read sequencing to accurately capture complex repeat regions (Chen et al, 2020) and the difficulties encountered in reconstructing sequence from closely related strains or sub–species (Bertrand et al, 2019; Quince et al, 2020; Vicedomini et al, 2021; Quince et al, 2017a). In practice a draft genome obtained from these methods would contain at best, tens and, more typically, hundreds, of distinct contigs, and so there are inherent difficulties in accurately determining the degree of genome completeness and the extent of contamination from non-cognate genomes (Chen et al, 2020), and in identifying the presence of horizontally transferred sequence (Douglas and Langille, 2019).…”