bcl2fastq_demux

WDL workflow to demultiplex Illumina NextSeq BCL data into paired-end gzipped FASTQ files.

Overview

Takes a tar.gz archive of an Illumina run directory and an Illumina-format SampleSheet.csv, and produces per-sample _R1.fastq.gz / _R2.fastq.gz file pairs.

Uses bcl_to_fastq (bcl2fastq-nextseq v1.3.0), which wraps Illumina's bcl2fastq2 and automatically concatenates reads across lanes.

Container: quay.io/biocontainers/bcl2fastq-nextseq:1.3.0--pyh5e36f6f_0

Inputs

| Parameter | Type | Default | Description | |---|---|---|---| | bcl_tar_gz | File | required | tar.gz of the Illumina run directory | | sample_sheet | File | required | Illumina SampleSheet.csv | | reverse_complement | Boolean | true | Reverse-complement index 2 (required for most NextSeq dual-index libraries) | | barcode_mismatches | Int | 0 | Allowed mismatches per index | | loading_threads | Int | 4 | Threads for BCL loading | | demux_threads | Int | 4 | Threads for demultiplexing | | processing_threads | Int | 8 | Threads for FASTQ processing | | writing_threads | Int | 4 | Threads for FASTQ writing | | memory_gb | Int | 16 | RAM in GiB | | disk_gb | Int | 500 | Disk in GiB (must exceed uncompressed BCL size) | | preemptible | Int | 1 | Preemptible VM retries (Google Cloud) |

Outputs

| Output | Type | Description | |---|---|---| | fastq_r1 | Array[File] | R1 FASTQ files, one per sample | | fastq_r2 | Array[File] | R2 FASTQ files, one per sample | | demux_stats | File | Per-sample read counts (demultiplexing_stats.csv) | | bcl2fastq_log | File | Raw bcl2fastq conversion log |

Usage

Terra.bio

1. Import the workflow from Dockstore into your Terra workspace.
2. Upload your run archive and sample sheet to a Google Cloud Storage bucket.
3. Fill in the inputs JSON (see bcl2fastq_demux_inputs.json for a template).
4. Launch the workflow.

Cromwell (local)

java -jar cromwell.jar run bcl2fastq_demux.wdl --inputs bcl2fastq_demux_inputs.json

Notes

• The tool uses --no-wait so it does not poll for RTAComplete; it assumes your tar.gz is a completed run.
• Undetermined FASTQ files are excluded from outputs by design.
• For single-index libraries, set reverse_complement to false.
• disk_gb should be at least 3× the size of the compressed tar.gz to accommodate extraction and FASTQ output.