RVI TOOLBOX

This reprosiroty provides reusable Nextflow workflows and submodules for the RVI’s pipeline ecosystem.

PREREQUISITES

• Nextflow (≥22.04.0)
• iRODS client (baton-do)
• Trimmomatic, TRF and sra-human-scrubber

IRODS WORKFLOW

The combined_input.nf and irods.nf provide workflows to fetch data from IRODS.

COMBINED_INPUT.nf

The combined_input.nf workflow consolidates and validates input data from two sources:

• Local Samplesheets: Validates FASTQ paths (via INPUT_CHECK).
• iRODS Metadata: Fetches sequencing data from iRODS using study/run/lane IDs (via IRODS_MANIFEST_PARSE and IRODS_CLI).

IRODS.nf

The irods.nf workflow fetches and processes sequencing data from iRODS, converting CRAMs to FASTQs while preserving metadata. It consists of two main sub-workflows:

1. QUERY_IRODS_METADATA

   Purpose: Queries iRODS for metadata (study, run, lane, etc.) using baton-do and parses JSON results.

   Key Steps:

   • Generates iRODS queries from input parameters (studyid, runid) or manifests.
   • Filters results (e.g., skips subsets like phix via params.irods_subset_to_skip).
   • Outputs tuples of (metadata, cram_path).

2. CRAM_EXTRACT

   Purpose: Retrieves CRAMs from iRODS, converts to FASTQs, and combines subsets (e.g., target + phix = total).

   Key Steps:

   • Skip Existing Files: Checks params.outdir for pre-existing FASTQs.

   • CRAM Conversion: Uses samtools collate (via COLLATE_FASTQ) to generate paired FASTQs.

   • Combine Subsets: Merges FASTQs from multiple subsets (e.g., target + phix) if combine_same_id_crams=true.

   • Cleanup: Removes intermediate files (cleanup_intermediate_files_irods_extractor=true).

PREPROCESSING Workflow

The PREPROCESSING.nf workflow processes raw FASTQ files through a modular pipeline:

Key Steps:

• Trimmomatic: Removes adapters (configurable via adapter_fasta parameter).
• Tandem Repeat Finder (TRF): Masks tandem repeats by converting FASTQ to FASTA, running TRF, and filtering repeats from the original reads.
• Human Read Removal: Uses sra-human-scrubber to exclude human-derived sequences.

Each step is toggleable via parameters (run_trimmomatic, run_trf, run_hrr). Inputs are paired-end FASTQs (provided as channel of the following structure tuples (meta, [read1, read2]), and outputs are cleaned FASTQs in the same format. The workflow validates paths (e.g., adapter files) and warns if all modules are disabled.

To run the workflow in isolation, use the following command:

nextflow run PREPROCESSING.nf -c PREPROCESSING.config \
  --preprocessing_mnf samples.csv \
  --adapter_fasta adapters.fa \
  --run_trimmomatic true

the csv (preprocessing_mnf) must have the following columns: sample_id,reads_1,reads_2.