WHALE: (W)orkflow for (H)uman-genome (A)nalysis of (L)ong-read (E)xperiments

Introduction

WHALE is a bioinformatics pipeline based on Nextflow and nf-core for long-read DNA sequencing analysis. It is optimized for the study and diagnosis of rare diseases. It takes a samplesheet as input and performs quality control, alignment, variant calling, variant integration and annotation.

Pipeline summary

1. Read QC (FastQC)
2. Present QC for raw reads (MultiQC)
3. Alignment (Minimap2)
4. Variant calling
   • Single Nucleotide Variant (SNV) calling (DeepVariant, Clair3, NanoCaller)
   • Structural Variant (SV) calling (Sniffles2, CuteSV, SVIM)
5. Merge variant calling
6. Annotation
   • SNV annotation (VEP)
   • SV annotation (AnnotSV)

Usage

First, prepare a samplesheet with your input data. Depending on which step of the analysis you want to run, the input data type can be: fastq, bam (and bai), vcf or bed. The samplesheet should look as follows:

samplesheet.csv:

sample,fastq
A123,/path/to/your/input/file/A123.fastq.gz
B456,/path/to/your/input/file/B456.fastq.gz

There are two types of full analysis:

• SNV analysis: -profile snv_analysis

• SV analysis: -profile sv_analysis

  Each full analysis can start with:

  • Alignment: --step mapping (input data type: fastq) (default)
  • Variant calling: --step variant_calling (input data type: bam and bai)

A specific step of the analysis can be executed:

• SNV calling (and merge): -profile snv_calling (input data type: bam and bai)
• SV calling (and merge): -profile sv_calling (input data type: bam and bai)
• SNV annotation: -profile snv_annotation (input data type: vcf)
• SV annotation: -profile sv_annotation (input data type: bed)

Profiles to use in the CCC (UAM):

• -profile uam,singularity,batch
• -profile uam_allcontigs,singularity,batch

Profiles to use in the server:

• -profile tblabserver,singularity
• -profile tblabserver_allcontigs,singularity

Examples

SNV and SV analysis starting with variant calling in the server:

nextflow run WHALE \
   -profile snv_analysis,sv_analysis,tblabserver,singularity \
   --input samplesheet.csv \
   --outdir <OUTDIR> \
   --step variant_calling

SV calling in the CCC:

nextflow run WHALE \
   -profile sv_calling,uam,singularity,batch \
   --input samplesheet.csv \
   --outdir <OUTDIR>

Pipeline output

WHALE will create the following subdirectories in the output directory:

• alignment
• snv_calling
  • snv_merge
• snv_annotation
• sv_calling
  • sv_merge
• sv_annotation
  • overlapping_sv_samples
• multiqc
• pipeline_info

Citations

An extensive list of references for the tools used by the pipeline can be found in the CITATIONS.md file.

Illustration by Yolanda Benítez