Nextflow Workflows

Data workflows for germline joint variant calling, somatic tumor-only and tumor-normal variant calling, and panel of normals creation.

Overview

This pipeline is designed to run germline joint variant calling and somatic variant calling from either *.fastq or *.bam input files using GATK best practices. It was developed to handle a variety of common workflow considerations, such as custom intervals for WES/WGS, inclusion of paired normals, creation of a panel of normals, and quality control reporting and metrics. The main subworkflows can be seen in the diagram below.

Getting Started

To run these workflows, you will need to install Nextflow (23.04.3) and Singularity (3.8.7) on an x86_64 machine. On the first run, a Docker image will be pulled to create a Singularity container with all of the required software for the workflow.

 _   _           _    __ _                        
| \ | | _____  _| |_ / _| | _____      __         
|  \| |/ _ \ \/ / __| |_| |/ _ \ \ /\ / /         
| |\  |  __/>  <| |_|  _| | (_) \ V  V /          
|_| \_|\___/_/\_\\__|_| |_|\___/ \_/\_/           
\ \      / /__  _ __| | __/ _| | _____      _____ 
 \ \ /\ / / _ \| '__| |/ / |_| |/ _ \ \ /\ / / __|
  \ V  V / (_) | |  |   <|  _| | (_) \ V  V /\__ \
   \_/\_/ \___/|_|  |_|\_\_| |_|\___/ \_/\_/ |___/

Samples:          null
Reference Genome: ${PWD}/reference/Homo_sapiens_assembly38.fasta

S U B W O R K F L O W   O P T I O N S
=====================================
--input=null
  <string> input file type (options: fastq, bam)

--subworkflow=null
  <string> subworkflow (options: joint-call, somatic-call, make-pon)

--out=OUT
  <string> name of outputs directory

--pon=null
  <string> path to panel of normals; used in somatic-call subworkflow

--gendb=GENDB
  <string> name of new genomics database; used in joint-call and make-pon subworkflow

--updategendb=null
  <string> path to existing genomics database; used when adding a new sample to genomicsdb in joint-call or make-pon subworkflow

--paired=false
  <boolean> determines if somatic-caller subworkflow is run with a paired normal

Example Usage

Sample Sheet Format

All subworkflows require a sample sheet, which is a headerless csv with either two or three columns. Generally, each row provides a sample identifier and paths to fastq/bam files for the corresponding sample. If starting from paired-end short read *.fastq inputs:

sample01,/absolute/path/to/sample01-R1.fastq.gz,/absolute/path/to/sample01-R2.fastq.gz
sample02,/absolute/path/to/sample02-R1.fastq.gz,/absolute/path/to/sample02-R2.fastq.gz
sample03,/absolute/path/to/sample03-R1.fastq.gz,/absolute/path/to/sample03-R2.fastq.gz

If you are starting from aligned, analysis-ready *.bam files:

sample01,/absolute/path/to/sample01.bam
sample02,/absolute/path/to/sample02.bam
sample03,/absolute/path/to/sample03.bam

Alternatively, if you want to run the somatic call workflow with a paired normal, you must use a three-columns sample sheet, where the second and third columns are paths to the tumor and normal *.bam files, respectively.

sample01,/absolute/path/to/sample01-tumor.bam,/absolute/path/to/sample01-normal.bam
sample02,/absolute/path/to/sample02-tumor.bam,/absolute/path/to/sample02-normal.bam
sample03,/absolute/path/to/sample03-tumor.bam,/absolute/path/to/sample03-normal.bam

FASTQ Inputs

# FASTQ to BAM
nextflow run --input fastq --samples /path/to/sample-sheet.csv main.nf

# FASTQ to Germline Cohort VCF
nextflow run --input fastq \
  --subworkflow joint-call \
  --samples /path/to/sample-sheet.csv \
  main.nf

# FASTQ to Tumor-Only VCF
nextflow run --input fastq \
  --subworkflow somatic-call \
  --samples /path/to/sample-sheet.csv \
  main.nf

BAM Inputs

# BAM to Germline Cohort VCF
nextflow run --input bam \
  --subworkflow joint-call \
  --samples /path/to/sample-sheet.csv \
  main.nf

# BAM to Tumor-Only VCF
nextflow run --input bam \
  --subworkflow somatic-call \
  --samples /path/to/sample-sheet.csv \
  main.nf

# BAM to Tumor-Normal VCF
nextflow run --input bam \
  --paired \
  --subworkflow somatic-call \
  --samples /path/to/sample-sheet.csv \
  main.nf

Updating an Existing Genomics Database

If you want to use an existing genomics database for new samples, provide the absolute path to the genomics database using the --updategendb parameter. NB: the name of all genomics database directories is set as GENDB, but you can change this using the --gendb parameter. Below is an example of adding new *.fastq samples to an existing genomics database and joint-calling the updated cohort.

nextflow run \
  --input fastq \
  --subworkflow joint-call \
  --samples /path/to/sample-sheet.csv \
  --updategendb /absolute/path/to/genomics-databse \
  main.nf

Outputs

The output folder has a default folder name OUT, however it can be renamed with the --out parameter. The exact contents of OUT will depend on your subworkflow.

OUT/
├─ bams/
│  ├─ <sample01>-markdup-bqsr.bam
│  ├─ <sample02>-markdup-bqsr.bam
│  ├─ ...
├─ gvcfs/
│  ├─ <sample02>.g.vcf.gz
│  ├─ <sample01>.g.vcf.gz
│  ├─ ...
├─ <genomicsdb_name>/
├─ cohort-vcfs/
│  ├─ cohort-<genomicsdb_name>.vcf.gz
│  ├─ cohort-<genomicsdb_name>-snp-recal.vcf.gz
│  ├─ cohort-<genomicsdb_name>-snp-indel-recal.vcf.gz
├─ pon/
│  ├─ pon-<genomicsdb_name>.vcf.gz
├─ mutect2/
│  ├─ filtered/
│  │  ├─ <sample01>-mutect2-filtered.vcf.gz
│  │  ├─ ...
│  ├─ unfiltered/
│  │  ├─ <sample01>-mutect2.vcf.gz
│  │  ├─ ...
├─ reports/
│  ├─ <tool-name>-<sample01>.html
│  ├─ <tool-name>-<sample02>.html
│  ├─ ...
├─ logs/
│  ├─ <tool_name>-<sample01>.log
│  ├─ <tool_name>-<sample02>.log
│  ├─ ...