Demo pipeline for annotating variants in .vcf files using Variant Effect Predictor (VEP).
Clone this repo:
git clone https://github.com/stevekm/vep-annotation-nf.git
cd vep-annotation-nf
Install nextflow
in the current directory with the command in the Makefile.
make install
To install VEP using Docker, run the Makefile command in the container
directory.
cd container
make docker-build
To install VEP using conda
(for NYULMC Big Purple HPC), instead run the conda-install
recipe from the Makefile in the parent repo directory.
make conda-install
VEP reference files will be downloaded automatically by the pipeline. However the hg19 genome fasta, fasta.fai, and fasta.dict files must also be obtained (not included; try these). On NYULMC Big Purple, all required files are already cached and no download should be needed. On other systems, the command line arguments specifying the genome fasta files should be provided separately when running, or place the files genome.fa
, genome.fa.fai
, and genome.dict
inside the included ref
directory.
The Makefile includes shortcuts to help run the pipeline easier on NYULMC Big Purple HPC.
make run
The command can also be used to run on other systems, it will simply invoke the command:
./nextflow run main.nf -resume
Nextflow params
values can be passed on the command line:
./nextflow run main.nf -resume --ref_fa /path/to/genome.fa --ref_fai /path/to/genome.fa.fai --ref_dict /path/to/genome.dict
Output files will be collected in the output
directory.
Tested on RHEL 7, macOS 10.12
-
Nextflow (Java 8+)
-
bash
-
GNU
make
-
Python 2.7+