Pipeline for processing multiple fastq files and running UMIerrorcorrect.

• UMIErrorCorrect
• bwa
• (optional) fastp

The tools can be installed from pip or conda, respectively:

 conda install -c bioconda bwa
 conda install -c bioconda fastp
 pip install umierrorcorrect

• one or more fastq files
• reference genome indexed with bwa
• (recommended) bed file containing amplicon regions

For basic usage only a directory containing fastq files and an indexed reference genome need to be supplied. A bed file containing amplicon annotations is optional, but recommended.

By default the pipeline assumes that fastq files are present as paired end files, i.e. two files per sample with one file containing "R1" and the corresponding "R2" in their file name.

umi_pipeline.sh -i <path_to_fastq_dir> -r <reference_fasta> -b <optional_bed_file>

If fastp is installed (recommended) it will automatically be used to trim adapters and perform quality filtering. Fastp will also merge R1 and R2 and perform error correction on the overlap region. An html report will be generated for each input fastq.

A number of optional flags can be set to modify the behaviour of fastp:

 -f  Do not perform filtering
 -q  Phred score threshold for filtering, default is 20.
 -p  Percent of reads that are allowed to have poor quality. Default is 40 (0-100).

For each fastq/sample a folder will be generated into which umierrorcorrect outputs will be redirected. Similarl to fastp, the arguments passed to UMIerrorocorrect can also be adjusted:

 -u <UMI length> Integer, default is 19.
 -s <Spacer length> Integer, default is 16.
 -t <Threads> Integer, default is 16.