Awesome

purple-nf

Nextflow pipeline for CNV calling with PURPLE

Description

Pipeline using PURPLE for copy number calling from tumor/normal or tumor-only sequencing data.

Usage

#using a tn_pairs file
nextflow run iarcbioinfo/purple-nf -r v1.0 \
-profile singularity  --tn_file tn_pairs..txt \
--cohort_dir $PWD/CRAM \
--ref hs38DH.fa --ref_dict hs38DH.dict \
--output_folder PURPLE


#activate BAM files mode
nextflow run iarcbioinfo/purple-nf -r v1.0 \
-profile singularity  --tn_file tn_pairs..txt \
--cohort_dir $PWD/BAM \
--bam \
--ref hs38DH.fa --ref_dict hs38DH.dict \
--output_folder PURPLE

Dependencies

1. This pipeline is based on nextflow. As we have several nextflow pipelines, we have centralized the common information in the IARC-nf repository. Please read it carefully as it contains essential information for the installation, basic usage and configuration of nextflow and our pipelines.
2. External software:
   • PURPLE
   • COBALT
   • AMBER

You can avoid installing all the external software by only installing Docker or singularity. See the IARC-nf repository for more information.

Input (mandatory)

Example of Tumor/Normal pairs file (--tn_file)

A text file tabular separated, with the following header:

tumor_id	sample	tumor	normal
sample1_T1	sample1	sample1_T.cram	sample1_N.cram
sample2_T1	sample2	sample2_T.cram	sample2_N.cram
sample3_T1	sample3	sample3_T.cram	sample3_N.cram

Optional parameters

Output

results
├── AMBER                               # AMBER result directory
│   ├── S00016_T_AMBER
│   │   ├── amber.version
│   │   ├── S00016_T_N.amber.snp.vcf.gz
│   │   ├── S00016_T_N.amber.snp.vcf.gz.tbi
│   │   ├── S00016_T_T.amber.baf.pcf
│   │   ├── S00016_T_T.amber.baf.tsv
│   │   ├── S00016_T_T.amber.baf.vcf.gz
│   │   ├── S00016_T_T.amber.baf.vcf.gz.tbi
│   │   ├── S00016_T_T.amber.contamination.tsv
│   │   ├── S00016_T_T.amber.contamination.vcf.gz
│   │   ├── S00016_T_T.amber.contamination.vcf.gz.tbi
│   │   └── S00016_T_T.amber.qc
├── COBALT									# COBALT result directory	
│   ├── S00016_T_COBALT
│   │   ├── cobalt.version
│   │   ├── S00016_T_N.cobalt.gc.median.tsv
│   │   ├── S00016_T_N.cobalt.ratio.median.tsv
│   │   ├── S00016_T_N.cobalt.ratio.pcf
│   │   ├── S00016_T_T.chr.len
│   │   ├── S00016_T_T.cobalt.gc.median.tsv
│   │   ├── S00016_T_T.cobalt.ratio.pcf
│   │   └── S00016_T_T.cobalt.ratio.tsv
│   ├── .....
├── PURPLE									# PURPLE result directory	
│   ├── S00016_T_PURPLE
│   │   ├── circos							# CIRCOS directory with files for plotting
│   │   ├── purple.version
│   │   ├── S00016_T_T.purple.cnv.gene.tsv
│   │   ├── S00016_T_T.purple.cnv.germline.tsv
│   │   ├── S00016_T_T.purple.cnv.somatic.tsv.           # Somatic Copy Number Segments
│   │   ├── S00016_T_T.purple.purity.range.tsv
│   │   ├── S00016_T_T.purple.purity.tsv
│   │   ├── S00016_T_T.purple.qc
│   │   ├── S00016_T_T.purple.segment.tsv
│   │   └── S00016_T_T.purple.somatic.clonality.tsv
│   ├── .....    
└── purple_summary.txt     # Summary file for all tumors
├── nf-pipeline_info		# Nextflow information directory
│   ├── purple_dag.html
│   ├── purple_report.html
│   ├── purple_timeline.html
│   ├── purple_trace.txt
│   └── run_parameters_report.txt # Custom file providing info for software versions and calling parameters

Common errors

Singularity

The first time that the container is built from the docker image, the TMPDIR should be defined in a non parallel file-system, you can set this like:

export TMPDIR=/tmp

multiple input files for each of the following file names: null, null.bai (or null, null.crai if using cram mode)

If the input file is not tabulation-separated, nextflow will return an input file name collision because it will not find the proper tumor and normal columns and thus will try to create dead symlinks to two sets of files named "null" and "null.bai" (see issue https://github.com/IARCbioinfo/purple-nf/issues/2 ):

Error executing process > 'AMBER'

Caused by:
  Process `AMBER` input file name collision -- There are multiple input files for each of the following file names: null, null.bai

The solution is to format the input file as a tsv, replacing space separations by tabulations.

Contributions