Yet Another Amplicon denoising Pipeline (YAAP), is a pipeline to analyse metabarcoding amplicon data. It performs QC, adapter removal, denoising and ZOTU table construction.
This is a fork to tidy up the documentation and bash formatting so I could understand it better. Original repo at https://github.com/CristescuLab/YAAP .
This pipeline makes use of the following programs:
- cutadapt (https://cutadapt.readthedocs.io/en/stable/index.html)
- vsearch (https://github.com/torognes/vsearch)
- seqkit (https://bioinf.shenwei.me/seqkit/)
- pear (https://cme.h-its.org/exelixis/web/software/pear/)
- usearch (https://www.drive5.com/usearch/)
- fastqc (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/)
This git contains the binaries (executables) of pear, seqkit, and usearch, however, vsearch and cutadapt have to be installed locally.
- All the dependencies for this pipeline (except usearch) can be easily installed using conda.
# grab the conda environment specification file
wget 'https://raw.githubusercontent.com/ilevantis/YAAP/master/yaap_env.yml'
# install the specified dependencies with conda
conda env create -f yaap_env
-
Now install the appropriate version of usearch from https://www.drive5.com/usearch/ .
-
Test that all dependencies work:
cutadapt -h
vsearch -h
pear -h
seqkit -h
usearch
- Put a copy of ASV_pipeline.sh into your bin folder and make it executable:
cd ~/bin
wget 'https://raw.githubusercontent.com/ilevantis/YAAP/master/ASV_pipeline.sh'
chmod +x ASV_pipeline.sh
YAAP has a single bash script called ASV_pipeline.sh. It requires 8 arguments:
- File with a list of fileset prefix (one per line)
- Forward primer sequence
- Reverse primer sequence
- Primer name
- Prefix of the output
- Number of cpus to use
- Minimum amplicon length
- Maximum amplicon length
- Unoise minimum abundance parameter
This pipeline assumes that your files are names fileset_prefix_R1.fastq.gz and fileset_prefix_R2.fastq.gz for all the samples. It also assumes that you have demultiplexed your samples.
As an example, let's assume that we have two samples named MC1 and MC2. Your
demultiplexed fastq files are MI.M03555_0320.001.N701N517.MC1_R1.fastq.gz,
MI.M03555_0320.001.N701N517.MC1_R2.fastq.gz,
MI.M03555_0320.001.N702N517.MC2_R1.fastq.gz,
MI.M03555_0320.001.N702N517.MC2_R1.fastq.gz.
You will need to create a file that looks like this:
MI.M03555_0320.001.N701N517.MC1
MI.M03555_0320.001.N702N517.MC2
The args 2 and 3 are the forward and reverse primer sequences. Forward primer is written as expected (5'-3'). Reverse primer is also written in the 5'-3' orientation - i.e. what you would see at the beginning of a read derived from the reverse strand of the amplicon.
For the following dsDNA amplicon:
dsDNA:
5'-ACTAGgctgactgTTTAG-3'
3'-TGATCcgactgacAAATC-5'
read from forward strand:
ACTAGgctgactgTTTAG
read from reverse strand:
CTAAAcagtcagcCTAGT
Forward primer = ACTAG Reverse primer = CTAAA
Example command:
The sample fastq files are listed in file_list.txt. The primers used were
GGWACWGGWTGAACWGTWTAYCCYC as forward, and TAAACTTCAGGGTGACCAAAAAATCA as
reverse. Running on a machine with 28 cpus. You can run the pipeline as:
bash ~/YAAP/ASV_pipeline.sh file_list.txt \
GGWACWGGWTGAACWGTWTAYCCYC TAAACTTCAGGGTGACCAAAAAATCA \
COI test 28 293 333 4
This call of the pipeline will focus on reads that contained the primer
sequences, that were longer than 126 (we focused on reads that are longer than
(min_len/2) - 20) in either direction (forwards and reverse), and which
assembly shoud be between 293bp and 333 (around expected length of the Leray
amplicon). The denoising would filter out reads with lower abundances than 4.
It will create an output folder with the name of the primer (COI in
the example) and append test to the output files. In the example, the output
folder will be called test_outputCOI.
If you find any bugs related with the pipeline, please open an issue in the the github repo, and add some traceback (error) information.