This is a nextflow wrapper script for Picard's IlluminaBasecallsToSam. It is designed to work efficiently on a computing cluster, but will work on smaller machines as well.
Run inside a run directory, it demultiplexes the run and produces unaligned bam (uBam) files
nextflow run demux_illumina.nf \
--read_structure <read structure> \
--flowcell <flowcell name> \
--max_mismatches <maximum allowed mismatches> \
--min_mismatch_delta <minimum distance to next closest barcode> \
--max_no_calls <maximum allowed no calls (i.e., Ns)> \
--lanecount <number of lanes> \
--machine <sequencing machine> \
--sequencing_center <sequencing center> \
--path_to_java <path to java> \
--path_to_picard <path to picard>
You will need to modify the path to picard variable to fit your environment. This script is configured to use up to ~64G of RAM per lane demultiplexing job, and up to ~16G of RAM per tile bcl->uBam job. If your machine is smaller, set the Xmx value appropriately.
- Picard tools and Java runtime
- Nextflow
Patches and bug reports are welcome.