Live bacterial species typing with Oxford Nanopore sequencing reads
This is a collection of tools and scripts which are used to perform real-time bacterial species typing of Nanopore reads as they come off the instrument. It is built around the Japsa framework, and consists of the following:
- Set up an analysis server
- Stream data to the analysis server
- Read output from server and visualize on a web page
This is just a prototype, used to generate files needed for the dev of the web page.
First, start netcat to listen to a specific port:
nc -k -l 127.0.0.1 3456 > typing-test.tsv
On a machine with a lot of RAM, this pipeline streams the output of jsa.np.rtSpeciesTyping to the specified port:
minimap2 -a -x map-ont -t 6 -K 100000 --secondary=no \
~/db/SpeciesTyping/Bacteria/genomeDB.mmi data/*.fastq | \
jsa.np.rtSpeciesTyping -bam - -index ~/db/SpeciesTyping/Bacteria/speciesIndex --read 50 -time 10 -out - | \
jsa.util.streamClient --input - --server 127.0.0.1:3456
The minimap2 database used is constructed from the RefSeq bacterial genomes from NCBI, and is already indexed (minimap2 -x map-ont -d genomeDB.mmi) to speed up analysis. Still, it is quite big and a machine with at least 24 Gb RAM is needed.