SigScreen is a nextflow pipeline for end to end mutational signature analysis.

The current implementation leverages SigVerse packages and sigminer for signature fitting.

Ensure nextflow is installed, then run sigscreen:

Create a manifest file (tsv) with 4 columns:

1. sample (required) sample identifier

2. snv (optional) path to vcf file with SNVs, MNVs, and INDELs.

3. copynumber (optional) path to segment file describing copynumber changes. Must be parse-able by sigstart::parse_cnv_to_sigminer().

4. sv (optional) path to segment file describing structural variant changes. Must be parse-able by sigstart::parse_purple_sv_vcf_to_sigminer()

At least one of the mutation files must be supplied for each sample.

Run signature analysis for every sample:

nextflow run -profile docker selkamand/sigscreen \
    --manifest=/path/to/manifest.tsv \
    --ref=hg38 \
    --n_bootstraps=25 \
    --cores=1

You may need to replace -profile docker with -profile singularity if working in linux environment.

Clone then cd into the repo and build the image with docker

docker build --tag sigscreen:v0.0.1

From inside the sigscreen directory run:

docker run --rm -v ./testdata/:/app/testdata \
    selkamandcci/sigscreen:v0.0.3 ./sigscreen.R \
        --manifest=manifest.tsv
        --ref=hg38 \
        --output_dir=/app/testdata/signatures \
        --n_bootstraps=25 \
        --temp_dir=/app/temp \
        --cores=1

docker run -it -v ./testdata/:/app/testdata selkamandcci/sigscreen:v0.0.3 sh