SpliceAI Rocksdb

Fast look up interface and prediction for missing variants in look table for spliceAI with rocksdb.

Installation

Clone this git repository:

git clone https://github.com/gagneurlab/spliceai_rocksdb.git

Go to the cloned repository:

cd spliceai_rocksdb

Create conda environment:

# Recommended if you have mamba installed
mamba env create -f environment.yml
# otherwise
conda env create -f environment.yml

Activate conda environment:

conda activate spliceai-rocksdb

Install modules from spliceai_rocksdb

pip install -e .

Download database

Download rocksdb for gnomad

spliceai_rocksdb_download --version {version} --db_path {output_path}

Supported version (grch37, grch38)

Usage

To run spliceai_rocksdb in the command line use:

spliceai_rocksdb -I input.vcf -O output.[csv, parquet, vcf] -R genome.fa -A grch37 -db {db_path}

Alternatively use the snakemake pipeline in /workflow

cd workflow
# modify config.yaml and setup datasets
python -m snakemake -j 1