SoFuR: SOmatic FUsions from Rna

Table of Contents

• Description
• Workflow steps
• Installation
• Usage
• Outputs directory
• Performances
• Copyright
• Contact

Description

This workflow detects, annotates and filters somatic fusions from stranded paired-end RNA-seq from Illumina's instruments.

Workflow steps

Installation

1. Download code

  git clone [--branch ${VESRSION}] --recurse-submodules git@github.com:bialimed/sofur.git

2. Install dependencies

• conda (>=4.6.8):

  # Install conda
  wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh && \
  sh Miniconda3-latest-Linux-x86_64.sh
  
  # Install mamba
  conda activate base
  conda install -c conda-forge mamba
  

  More details on miniconda install here.

• snakemake (>=5.4.2):

  mamba create -c conda-forge -c bioconda -n sofur snakemake==7.32.4
  conda activate sofur
  pip install drmaa
  

  More details on snakemake install here.

• Install rules dependencies (cutadapt, bwa, ...):

  conda activate sofur
  snakemake \
    --use-conda \
    --conda-prefix ${application_env_dir} \
    --conda-create-envs-only
    --snakefile ${APP_DIR}/Snakefile \
    --configfile workflow_parameters.yml
  

4. Download and prepare resources

SoFuR uses genome sequences, genes annotations, known fusions databanks (artifacts and pathogenics) and others standard resources. These files must be provided in your config.yml (see template). An detailed example for the creation process of these resources is available here.

5. Test install

• From ${APP_DIR}/test/config/wf_config.yml set variables corresponding to databanks (see ## ${BANK}/... ##).

• Launch test wit following command:

  conda activate sofur
  ${APP_DIR}/test/launch_wf.sh \
    ${CONDA_ENVS_DIR} \
    ${WORK_DIR} \
    ${DRMAA_PARAMS}
  

  Example with scheduler SGE:

  export DRMAA_LIBRARY_PATH=${SGE_ROOT}/lib/linux-rhel7-x64/libdrmaa.so
  
  conda activate sofur
  ~/soft/sofur/test/launch_wf.sh \
    /work/$USER/conda_envs/envs \
    /work/$USER/test_sofur \
    ' -V -q {cluster.queue} -l pri_{cluster.queue}=1 -l mem={cluster.vmem} -l h_vmem={cluster.vmem} -pe smp {cluster.threads}'
  

  Example with scheduler slurm:

  export DRMAA_LIBRARY_PATH=$SGE_ROOT/lib/linux-rhel7-x64/libdrmaa.so
  
  conda activate sofur
  ~/soft/sofur/test/launch_wf.sh \
    /work/$USER/conda_envs/envs \
    /work/$USER/test_sofur \
    ' --partition={cluster.queue} --mem={cluster.mem} --cpus-per-task={cluster.threads}'
  

• See results in ${WORK_DIR}/report/run.html.

Usage

Copy ${APP_DIR}/config/workflow_parameters.tpl.yml in your current directory and change values before launching the following command:

conda activate sofur
snakemake \
  --use-conda \
  --conda-prefix ${application_env_dir} \
  --jobs ${nb_jobs} \
  --jobname "sofur.{rule}.{jobid}" \
  --latency-wait 100 \
  --snakefile ${application_dir}/Snakefile \
  --cluster-config ${application_dir}/config/cluster.json \
  --configfile workflow_parameters.yml \
  --directory ${out_dir} \
  > ${out_dir}/wf_log.txt \
  2> ${out_dir}/wf_stderr.txt

Outputs directory

The main elements of the outputs directory are the following:

out_dir/
├── ...
├── report/
|   ├── ...
|   ├── run.html
|   └── sample-A.html
├── stats/
|   ├── ...
|   └── multiqc/
|       ├── ...
|       └── multiqc_report.html
└── structural_variants/
    ├── ...
    ├── sample-A_unfiltered.vcf
    └── sample-A_filtered.vcf

• The reports files containing filtered fusions list, annotations and viewers are in out_dir/report/{sample}.html.
• The quality reports is in out_dir/stats/multiqc/multiqc_report.html. It resumes qualities of reads, distribution of alignments (between exon, intron, ...) and strandness analysis.
• The annotated fusions in VCF format are kept in out_dir/structural_variants. *_unfiltered.vcf contain all the fusions, their annotations and their tags. *_filtered.vcf contain all the fusions, their annotations and their tags after filtering by rules provided by file declarated in configfile by filters.rules.

Performances

The performances are evaluated on real, synthetic and simulated datasets. The commands used in evaluation are stored in assessment. The results summarized in assessment/report.html.

Copyright

2019 Laboratoire d'Anatomo-Cytopathologie du CHU de Toulouse