ampliconarchitect-nf

Nextflow pipeline to discover ecDNA in cancer genomes.

Usage

  #using a tn_pairs file
  nextflow run iarcbioinfo/ampliconarchitect-nf -r v1.0 \ 
  -profile singularity  --tumor_file tumor_file.txt \
  --cram_dir $PWD/CRAM \
  --moseklm_license mosek.lic --aa_repo_dir GRCh38 \
  --output_folder ecDNA_public1
  
  #testing a small fake run
   nextflow run iarcbioinfo/ampliconarchitect-nf -r v1.0 \ 
   --tumor_file test/tumor_file.txt \
   --cram_dir test/cram \
   --moseklm_license test/mosic_license/mosic.license \
   --aa_repo_dir test/amplicon_repo --debug true

Dependencies

License for Mosek optimization tool

Obtain license file mosek.lic (https://www.mosek.com/products/academic-licenses/ or https://www.mosek.com/try/).

Amplicon architect repository

To run amplicon architect is necessary to download the Amplicon Architect repository from here, currently we have tested this nextflow pipeline with GRCh38 genome built.

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

Input

Mandatory

Parameter	Description
--cram_dir	Directory containing all BAM/CRAM files
--tumor_file	File containing list of tumor CRAM files to be processed
--moseklm_license	License for mosek program [mosic.license]
--aa_repo_dir	directory containing the Amplicon Architect repository [See above]

Optional

Parameter	Description
--output_folder	name of output folder [def:amplicon_results]
--cpu	Number of CPUs[def:1]
--mem	Max memory [def:8Gb]
--debug	enable debuging of the pipeline

Example of Tumor file (--tumor_file)

A text file tabular separated, with the following header:

id	cram
sample1_T1	sample1_T.cram
sample2_T1	sample2_T.cram
sample3_T1	sample3_T.cram

Output

├── amplicon_classes                # amplicon classified
│   └── all_amplicons_classification_bed_files # amplicon in genomic coordiantes 
├── amplicon_predictions           # Amplicon architect predictions
│   ├── S00016_T.aa
│   ├── S01493_T.aa
│   ├── S01501_T.aa
├── amplified_intervals          #Selected seeds per sample
├── cnvkit							#CNVkit copyNumber segments
│   ├── S00016_T
│   ├── S01493_T
│   ├── S01501_T
│   ├── S01502_T
└── nf-pipeline_info           # Nextflow logs

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

Contributions

Name	Email	Description
Matthieu Foll*	follm@iarc.fr	Developer to contact for support (link to specific gitter chatroom)
Alex Di Genova	digenovaa@fellows.iarc.fr	Developer