nRex

nRex is a set of script to call germline and somatic single-nucleotide variants, short indels and structural variants. The workflow is primarily for human whole-genome sequencing data (>=30x) using short or long-reads.

Installing nRex

git clone --recursive https://github.com/tobiasrausch/nRex.git

cd nRex/

make all

Download reference genome

cd genome/ && ./prepare_genome.sh

Download a haplotype reference panel

cd refpanel/ && ./download.sh

cd maps/ && ./download.sh

Optional: VEP annotation of variants

To activate the annotation of variants using VEP, you need to download an offline annotation cache for GRCh38

make .vep

Running nRex for GRCh38

There is a pipeline for short- and long-reads. For short-reads:

./src/nRex.sh <output prefix> <sample1.read1.fq.gz> <sample1.read2.fq.gz>

and for long-reads:

./src/nRexLR.sh <output prefix> <sample.fq.gz>

Postprocessing the output of the pipeline

A few helper scripts to summarize the output of the various tools.

Aggregating QC statistics across multiple samples

./src/aggregate.sh table *.qc.summary

Summarizing VEP output

To generate a table of annotated variants, you can use

python3 scripts/vep.py -v sample.vep.bcf