ReVERSe

Rare Variant Enrichment and Recessive Segregation

Installation

Method 1:

ReVERSe requires several modules from VASE (https://github.com/david-a-parry/vase). First install VASE using pip (or for other methods look at the instructions in the README at the VASE github repo) and then install ReVERSe from either the gitlab or github repo.

python3 -m pip install git+git://github.com/david-a-parry/vase.git --user
python3 -m pip install git+https://git.ecdf.ed.ac.uk/dparry/reverse.git  --user

The --user flag is optional, but circumvents permissions issues that may arise depending on how your python libraries are configured.

Method 2:

First clone the repo and install with dependancies.

git clone https://git.ecdf.ed.ac.uk/dparry/reverse.git
cd reverse
python3 -m pip install . --process-dependency-links --user

If you get an error "no such option: --process-dependency-links" (with newer versions of pip) with the third command see method 3.

Method 3:

Note that the first two steps are not necssary if you've already run them while attempting Method 2.

git clone https://git.ecdf.ed.ac.uk/dparry/reverse.git
cd reverse
python3 -m pip install -r requirements.txt --user
python3 -m pip install . --user

Test

If you want to test your installation run the test.sh shell script in the test directory. This should complete without errors and produce a CSV output file in test/test_output.

Synopsis

# Get counts from gnomAD data - make sure you are using the same genome 
# build as the gnomAD VCF!

ReVERSe_count -i cohort.vcf.gz \
    --ped cohort.ped \
    -e gnomad.exomes.r2.1.sites.vcf.gz \
    -g gnomad.genomes.r2.1.1.sites.vcf.bgz \
    --exome_coverage_files gnomad.exomes.coverage.summary.tsv.bgz \
    --genome_coverage_files gnomad.genomes.coverage.summary.tsv.bgz \
    --gnomad_version 2.1 \
    -v cohort_rev_counts.vcf.gz \
    -t cohort.rev_counts.txt.gz 

# Perform recessive segregation analysis requiring either biallelic
# enriched alleles or enriched alleles in trans with VEP HIGH impact
# variants, looking at variants with a maximum frequency of 0.5 %. Alleles
# with a Fisher's exact test P-value <= 1e-5 are considered enriched.

ReVERSe_seg -i cohort_rev_counts.vcf.gz \
    --ped cohort.ped \
    --freq 0.005 \
    --canonical \
    --pops afr amr eas fin nfe sas \
    -v 1e-5 \
    --impact HIGH \
    -o cohort_rev_seg.vcf.gz

# Output ranked variants in CSV format

ReVERSe_reporter cohort_rev_seg.vcf.gz \
    cohort_rev_seg.report.csv \
    cohort.ped

Note that each of the programs have many different options that can be provided to tweak parameters. For detailed help run the desired program with the --help flag.

gnomAD VCFs and coverage data are available from https://gnomad.broadinstitute.org/downloads#v2-coverage