Some useful resource to give a try

• https://www.researchgate.net/post/Any_bioinformatics_tool_method_to_identify_plasmid_sequences_from_bacterial_whole_genome_data
• https://ccb-microbe.cs.uni-saarland.de/plsdb/plasmids/search_form/seq/
• http://www.patlas.site/
• you can also use native blast or diamond together with plasmid database for the custom analysis with your own parameters.

Dear Tanjian,

This is an interesting problem, regardless of whether you know a plasmid is present in the strain that was sequenced or not. One simple solution would be to search for plasmid replication, partition and transfer functions, and work out from there. It is likely that the plasmid will occupy a single contig, but if you had a relatively low coverage, it may appear in smaller contigs which are divided buy repeat elements. You could also scan G+C content and look at any areas with unusually high or low G+C%, as it is often that plasmids differ from the host genome composition.

Regards, Andrew

Dear Tangjian

you can assemble plasmid DNA short reads from bacterial whole genome data using plasmidSpades tool.

Good luck,

Marcela