umi_nextflow

Developmental pipeline for high-sensitivity variant calling using UMI-tagged short sequence reads.

Dependencies

• Nextflow: wget -qO- https://get.nextflow.io | bash
• Docker

Running the pipeline

The pipeline is initiated from a directory containing paired-end fastqs. These fastqs should be of the format <sample_id>.{1,2}.fastq.gz

One can intiate the pipeline using a command structured like so:

nextflow run main.nf -profile local --input_dir 302R_fastqs/ --run_id 302R --downsample_reads 25000000

Nextflow will create the directory work for scratch work and analysis. By default, the output from the pipeline will be saved to the output directory.

If the pipeline were to encounter an error, one that error is addressed, the pipeline can be resumed from a cached state by appending the -resume flag to the original command. Resuming the above example from a cached state would look like:

nextflow run main.nf -profile local --input_dir 302R_fastqs/ --run_id 302R --downsample_reads 25000000 -resume

Parameters

Required

• -profile The config profile for the run environment. Options {local, uw_batch}
• --input_folder Path to a directory containing paired-end fastqs. Trailing / is required
• --run_id An ID for the sequencing run

Optional

• --downsample_reads Number of reads to sample from each fastq. Currently using 25000000 for research runs
• --output Diretory in which to save pipeline output. Default './output'
• --save_intermediate_output Flag to save intermediate files to output directory
• -resume Flag to resume pipeline from a cached state, if possible

References

The mpileup2readcounts binary was derived from the IARC bioinformatics platform GitHub repository.