Table 2.

Summary of common in silico tools used for plasmid analysis.

Goal	Tool(s); reference(s)	Comments
Detect loci of interest from reads	SRST2 (Inouye et al., 2014)	Reads are mapped to a reference database using bowtie2 (Langmead and Salzberg, 2012). Some databases are included as part of the tool (e.g., PlasmidFinder, ResFinder, ARG-ANNOT) but custom databases can also be used.
Detect resistance genes from k-mers	KmerResistance (Clausen et al., 2016)	Identifies resistance genes from WGS data by examining co-occurrence of k-mers (DNA substrings of length k) between the query WGS data and a reference database of resistance genes.
Comparative plasmid genomics	ACT; BRIG (Carver et al., 2005; Alikhan et al., 2011)	Tools such as ACT and BRIG can be used to order contigs against a reference plasmid using BLAST, allowing homologies and gene content similarity to be visualized.
Detect replicon type/subtype from contigs	PlasmidFinder; pMLST (Carattoli et al., 2014)	See Table 1.
Detect resistance genes from contigs	ResFinder (Zankari et al., 2012)	Contigs are BLAST searched against a database of horizontally acquired resistance genes; resistance-conferring mutations are not accounted for.
	CARD (McArthur et al., 2013)	Contigs are BLAST searched against the CARD database; resistance genes are associated with an ontology allowing resistance gene metadata to be retrieved. CARD also provides the Resistance Gene Identifier tool for resistance prediction.
	ARG-ANNOT (Gupta et al., 2014)	BLAST-based tool for detection of resistance genes and resistance mutations.
Localize specific genes of interest from a contig assembly	Bandage (Wick et al., 2015)	Assembly graph visualization and annotation tool (can be used for manual repeat resolution).
	ISMapper (Hawkey et al., 2015)	Mapping-based tool which uses paired-end sequencing data to localize insertion sequences. Can be used for localizing a particular resistance locus, given a known association with a specific insertion sequence.
Distinguish plasmid from chromosomal sequences	cBar (Zhou and Xu, 2010)	Plasmid and chromosomal sequences are distinguished based on pentamer frequencies.
	Other tools	Tools such as plasmidSPAdes and PlasmidFinder may also be used to distinguish plasmid and chromosomal sequences (Arredondo-Alonso et al., 2016).
Resolve plasmid structures from ambiguous assembly graphs	PLACNET (Lanza et al., 2014)	An input assembly graph is reconfigured according to the homology of contigs to reference sequences; the assembly graph can be visualized to allow manual pruning and correction.
	Recycler (Rozov et al., 2016)	Cycles in an assembly graph are identified and sequentially extracted from the graph, favoring cycles with minimal coverage variation across constituent contigs. Assuming different genetic units have distinct copy numbers, retrieved cycles should represent individual circular elements (plasmids, circular phages). Information from paired-end reads is used to exclude cycles that do not correspond to a single circular element, but arise from repeat elements shared across different molecules.
	plasmidSPAdes (Antipov et al., 2016)	Median coverage of longer contigs is calculated to estimate chromosomal coverage; this estimate is used as a basis for filtering putative chromosomal contigs from the assembly graph. Connected components within the filtered graph are reported as putative plasmids. This approach assumes that chromosomal contig coverage differs from plasmid contig coverage.

The tools presented here are not necessarily exhaustive and not all are intended only for plasmid analysis.