The databases need to include "Plasmids" too

[tseemann]

Derrick,

I notice that often my pure bacterial samples still get 5% say of reads being unclassified. When I assemble these reads, they turn out to be bacterial plasmids.

The problem is that some of the Bacteria folders have chromosomes and plasmids, but there are also many separately submitted plasmids which are in a different Plasmids folder at NCBI:

ftp://ftp.ncbi.nih.gov/genomes/Plasmids/

It would be great to add support for this in the download tools, and in MiniKraken.

[DerrickWood]

Would the plasmids.all.fna.tar.gz file in that directory be sufficient for what you need? That would be easy enough to add to the download list.

[tseemann]

Yes, that would be sufficient. This will help explain the unaccounted for 10-20% of reads in some samples where the plasmids have high copy number. Thanks!

[DerrickWood]

OK, I've added the plasmids as an option. I've got to monitor a few things over the next few months to make sure that file is consistently available before I can make it part of the standard installation, but at least people can more easily bring it in. I'll also likely add it to at least some version of the MiniKraken database in the near future as well.

[tseemann]

Thank you.

The Plasmids folder has been there for many years, but not as long as Bacteria and Viruses, but NCBI do change things without notice. The annoying part is that Bacteria/ does contain some plasmids, just not all.

[nickp60]

Hi, sorry to revive an old issue, but just to clarify, did the plasmids end up in either of the MiniKraken databases?

[jenniferlu717]

The plasmids in the plasmids folder did not end up in the minikraken databases but I'll work to include them in another version ASAP. Might take a couple of days though.

[tseemann]

@jenniferlu717 what ended up happening here?