Improved detection and classification of plasmids from circularized and fragmented assemblies

Abstract

Plasmids are mobile genetic elements important for bacterial adaptation. The study of plasmids from sequencing data is challenging because short reads produce fragmented assemblies, requiring of subsequent discrimination between chromosome and plasmid sequences. Although circularized assemblies are now possible using long-read data, there is still a need to differentiate plasmids from other circular elements. Here, we present plaSquid, a dockerized tool developed in Nextflow that expands plasmid detection and improves replicon typing and mobility groups classification schemes, outperforming previously available methods in both precision and sensitivity. When applied to [~]10.5 million metagenomic contigs, plaSquid revealed a 2.7-fold increase in plasmid phylogenetic diversity. Also, we used plaSquid to uncover a significant role of plasmids in the widespread distribution of clinically-relevant antimicrobial resistance genes in the built environment, from cities to spacecraft. Together, we present an improved approach to study plasmid biology from fragmented or circularized genomic and metagenomic assemblies.