Remove vector sequences from genome database

Question

[Helena]

Hi,

I'm building a database containing Refseq genome sequences from selected bacterial species, which will be used for Nanopore sequencing of environmental samples.

In order to eliminate chances of false positives, I used the UniVec database to locate any potential contamination and got substantial hits to several vectors. I am pretty new to bioinformatics and therefore I wanted to hear if anyone has any ideas of how to mask/remove the contamination from the genome sequences?

/Helena

[GenoMax]

Are you only creating a database of main chromosomes from bacterial species? Normally the genomes may also include plasmids.

[Helena]

To start I'll create a database containing the chromosomes and afterwards I'll create one for plasmids :) I have already separated the plasmid sequences from the chromosomes.

Answer 1

[colindaven]

I also work on this, but for lung metagenomes chiefly. We have supplied ref seqs (unknown how useful for your purposes, as no plasmids) here: https://github.com/MHH-RCUG/Wochenende#installation

I wrote a contamination masking tool here, because otherwise it becomes a big source of false positives.

https://github.com/colindaven/blacklister