SOMvc pipeline

a variant calling pipeline from matched normal-tumor .bam files to .vcf files using somaticcombiner consensus approach utilizing lofreq, mutect2, strelka and vardict.

---

set up pipeline

before running, you have to set up the attached Docker images, DNAvc-pipeline docker image (for variant interpretation) can take some time:

docker build -t somvc-pipeline https://raw.githubusercontent.com/loipf/SOMvc-pipeline/master/docker/Dockerfile
docker build -t dnavc-pipeline https://raw.githubusercontent.com/loipf/DNAvc-pipeline/master/docker/Dockerfile

now either replace the Docker container hash (last output line from previous build command) in nextflow.config or run nextflow with the -with-docker somvc-pipeline argument.

no mapping-refinement is needed since VC-callers do so internally.

if you already have an reference genome and bed file, you can specify it with the nextflow argument --reference_genome [genome.fa.gz] and --bed_file [bed_file_filtered.bed] (only chromosomes which are available in the genome). otherwise run the following script to download it into data folder:

bash scripts/data_acquisition.sh

---

run mapping pipeline

it can be run locally with downloaded github-repo and edited nextflow.config file with:

nextflow run main.nf

or

nextflow run loipf/SOMvc-pipeline --project_dir /path/to/folder --sample_match_file normal_tumor_pairs.csv --reference_genome genome.fa.gz -resume --bed_file bed_file_filtered.bed --num_threads 10 -with-docker somvc-pipeline

for this execution to work properly, you have to be in the current project directory.

optional extendable with:

-resume
-with-report report_SOMvc-pipeline
-with-timeline timeline_SOMvc-pipeline
-w work_dir

by default, all output will be saved into the data folder of the current directory

input .bam files must be sorted with marked duplicates (and index created in the same directory).

the sample matching normal_tumor_pairs.csv file must be constructed like following (with same column names, filenames per sample must be different!):

sample_id,normal_file,tumor_file
sample_1,/path/to/sample_1.bam,/path/to/sample_1_tum.bam
sample_2,/path/to/sample_2.bam,/path/to/sample_2_tum.bam
sample_3,/path/to/sample_3.bam,/path/to/sample_3_tum.bam