The genetic and ecological landscape of plasmids in the human gut

Abstract

Plasmids play a critical role in rapid bacterial adaptation by encoding accessory functions that may increase the host's fitness. However, the diversity and ecology of plasmids is poorly understood due to computational and experimental challenges in plasmid identification. Here, we report the Plasmid Classification System (PCS), a machine learning classifier that recognizes plasmid sequences based on gene functions. To train PCS, we performed a large-scale discovery and comparison of gene functions in a reference set of >16,000 plasmids and >14,000 chromosomes. PCS accurately recognizes a diverse range of plasmid subtypes, and it outperforms the previous state-of-the-art approach based on k-mer decomposition of sequences. Armed with this model, we conducted, to our knowledge, the largest search for naturally occurring human gut plasmids in 406 publicly available metagenomes representing 5 countries. This search yielded 6,257 high-confidence predicted plasmids, of which 576 had evidence of a circular conformation based on pair-end mapping. These predicted plasmids were found to be highly prevalent across the metagenomes compared to the reference set of known plasmids, suggesting there is extensive and uncharacterized plasmid diversity in the human gut microbiome.

Diversity and ecology of Caudoviricetes phages with genome terminal repeats in fecal metagenomes from four Dutch cohorts

“…Notably, 20% of the taxonomically unassigned vOTUs represented by sequences with DTR were predicted to be plasmids with high confidence (PlasX score > 0.9 [32]), whereas the same was the case for only 2% of the Caudoviricetes vOTUs represented by sequences with DTR.…”

“…When considering virus-like genomes with DTR that did not receive any taxonomic assignment, it can be difficult to distinguish these from plasmids. Notably, 20% of the taxonomically unassigned vOTUs represented by sequences with DTR were predicted to be plasmids with high confidence (PlasX score > 0.9 [32]), whereas the same was the case for only 2% of the Caudoviricetes vOTUs represented by sequences with DTR. We therefore decided to focus on Caudoviricetes vOTUs represented by sequences with terminal repeats (Figure 1).…”

“…Nucleotide sequences with DTR characterized by a PlasX score > 0.9 were considered potential plasmids [32].…”

A low-cost genomics workflow enables isolate screening and strain-level analyses within microbiomes

“…S5). We enumerated all plasmids within our isolate assemblies using a recently developed machine learning approach [21] which leverages gene content to predict plasmids. A scan of all isolate assemblies yielded a total of 516 putative plasmid contigs recovered from 258 individual genome assemblies.…”

“…A scan of all isolate assemblies yielded a total of 516 putative plasmid contigs recovered from 258 individual genome assemblies. These belonged to 245 clusters as calculated by MobMess [21], with 64 of these clusters containing more than one contig. Only 31 of these putative plasmids matched existing plasmid sequences in the PLSDB plasmid database [22], indicating that more than 94% of the recovered plasmid sequences were novel.…”