Pipeline example: MiSeq 2x250 16S V4
OTU / denoising
Example pipelines with test
This example shows a typical analysis
pipeline for MiSeq paired reads. There are four samples: Human, Mouse, Soil
and Mock with ~4k reads each. Human and Mouse are fecal samples. Data is
from Kozich et al. 2013.
The mock reads have spurious OTUs due to cross-talk.
You need to set
the environment variable $usearch to the name (or path) for your
usearch binary file. The bash script
assumes that the FASTQ files and the sintax reference database are stored in
a directory called ../data. Output is written to a directory called ../out
which is deleted and re-build. I suggest making a project directory, say
ex_miseq, with sub-directories named data and scripts. Put the ex_miseq.bash
script in the scripts sub-directory and execute it from there.
Download commands and data
Reads: right-click on
and click Save As. Use tar -zxvf *.tar.gz to extract.
reference database: right-click on
rdp_16s_v16.fa.gz and click Save As. Use gunzip
Bash script: select all text below (ctrl+A) and
copy/paste into a text editor, or right-click on
ex_miseq.bash and click Save As.