Tool:Recycler for plasmid assembly

Question

[Roye Rozov]

Hi all,

I noticed there have been questions related to plasmid assembly popping up periodically, so I wanted to introduce you guys to our new tool, Recycler. Recycler takes assembly graphs output by recent versions of SPAdes and uses them to search for likely plasmids in isolate samples, metagenomes, and purified plasmidome data. The tool is freely available here: https://github.com/Shamir-Lab/Recycler

For those that want full details, the manuscript is here: http://biorxiv.org/content/early/2016/04/05/029926

Our tests and feedback from early users indicate Recycler's outputs are real, and we are conducting extensive biological validation of predicted plasmids. We welcome any feedback, and hope the tool helps advance the study of plasmids!

Cheers, RR

Answer 1

[Roye Rozov]

here is an update regarding the validation mentioned above: http://biorxiv.org/content/early/2016/05/09/029926

[Bioinformatics_NewComer]

Thanks for sharing this. I wonder how this is different from plasmidSPAdes:

http://biorxiv.org/content/early/2016/04/15/048942

Recycler uses files from SPAdes.

plasmidSPAdes is SPAdes' component providing desired plasmid assembly (I could be wrong here). They plan to ship with this option in next SPAdes release.

Thanks.

[Roye Rozov]

Thanks for the reply. I haven't investigated their method deeply, but here are the practical differences I have noticed: - Recycler looks to output one sequence per plasmid. If there is no complete cycle the sequence will not be reported. plasmidSPAdes in some cases will output a set of contigs that together may form a plasmid, instead of one sequence. - Recycler handles a more general case - it can be applied to assembly graphs of isolates, metagenomes, or plasmidomes. plasmidSPAdes is limited to isolates.

At some point we will also run a comparison, but biological validation was more urgent for us to confirm Recycler's predictions are real. Based on those results, we get real sequences nearly 90% of the time.

[Bioinformatics_NewComer]

Dear Roye,

Thanks much for the prompt reply, and sharing insights between the 2 tools. This definitely helps. I'm to go ahead with recycler for the time being. :)

Thanks again.