Somatic Driver Discovery Workflow

The somatic driver discovery workflow (SDD) identifies potential somatic drivers based on a user defined cohort. The inference is done via three methodologies: MutEnricher, oncodriveFML and dNdScv.

How to run the workflow

Step 1 - Set up the analysis folder

Create a folder to store your analysis results and copy the file submit.sh to this location.

mkdir /pgen_int_work/BRS/jdoe/path/to/analysis_folder
cd /pgen_int_work/BRS/jdoe/path/to/analysis_folder
cp /pgen_int_work/BRS/somatic_discovery/submit.sh .

Step 2 - Format the cohort file

Make a tab delimited file with your cohort of interest. This file should contain two columns and no header. The first column must have a VCF path while the second column the sample platekey. You can find an example here: /pgen_int_work/BRS/somatic_discovery/example/luad_lusc_example_samples.txt.

/path/to/VCF/sample_platekey_1.vcf.gz        sample_platekey_1
/path/to/VCF/sample_platekey_2.vcf.gz        sample_platekey_2
/path/to/VCF/sample_platekey_3.vcf.gz        sample_platekey_3
...

Step 3 - Update submission script

Edit the file submit.sh to point to the workflow folder, a scratch directory and the cohort file.

1. Set the workflow folder in line 18 DRIVER_DISCOVERY='/pgen_int_work/BRS/somatic_discovery'. If you have not copied the workflow folder somewhere else, this should be pre-populated with the right path.
2. Set the path to a scratch folder in line 19 SCRATCHDIR='/re_scratch/path/to/dir'
3. Provide the path to the cohort samples in line 24 --sample_file "/pgen_int_work/BRS/somatic_discovery/example/luad_lusc_example_samples.txt"

Here are the relevant lines in the submit.sh file:

# VARIABLES TO EDIT PRIOR TO SUBMISSION
DRIVER_DISCOVERY='/pgen_int_work/BRS/somatic_discovery'
SCRATCHDIR='/re_scratch/path/to/dir'

# RUN WORKFLOW
nextflow run "${DRIVER_DISCOVERY}"/main.nf \
    --variant_type "coding" \
    --sample_file "/pgen_int_work/BRS/somatic_discovery/example/luad_lusc_example_samples.txt" \
    --region_file "/pgen_int_work/BRS/somatic_discovery/resources/global/coding_CDS.tsv.gz" \
    --scratchdir "${SCRATCHDIR}" \
    -c "${DRIVER_DISCOVERY}"/nextflow.config \
    -profile cluster

Step 4 - Submit job to HPC cluster

Now, you only need to submit the analysis script to the cluster. You can do so from the analysis folder:

cd /pgen_int_work/BRS/jdoe/path/to/analysis_folder
bsub < submit.sh

How to read the results

The results for each tool are available in the results folder created inside the analysis directory.

Here's an example of what the results folder looks like:

cd /pgen_int_work/BRS/jdoe/path/to/analysis_folder

tree results/

results/
├── dndscv
│   └── dndscv
├── mutenricher
│   ├── test_gene_enrichments.txt
│   ├── test_gene_hotspot_Fisher_enrichments.txt
│   └── test_hotspot.txt
└── pipeline_info
    ├── execution_report_2024-07-11_15-02-15.html
    ├── execution_timeline_2024-07-11_15-02-15.html
    ├── execution_trace_2024-07-11_15-02-15.txt
    └── pipeline_dag_2024-07-11_15-02-15.html

Advanced use

If you'd like to customise a set of parameters, you can do so by editing the conf/tools_params.config file. However, you'll need to copy the workflow folder to an appropriate directory to be able to edit this file.

Running the workflow should be just as easy, provided you update the DRIVER_DISCOVERY variable in the submit.sh file with the new folder path.

Information on individual parameters can be found in the documentation for each tool, referenced below.

References

These resources provide more information on individual tools and methods applied to the workflow.

Tool	Source	Repository
dNdScv	http://www.cell.com/cell/fulltext/S0092-8674(17)31136-4	https://github.com/im3sanger/dndscv
MutEnricher	https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-020-03695-z	https://github.com/asoltis/MutEnricher
oncodriveFML	https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016-0994-0	https://bitbucket.org/bbglab/oncodrivefml/src/master/