VEP Cache Extractor

Introduction

The VEP cache extractor is a command line utility for extracting fields from the cache used by Ensembl's VEP, or Variant Effect Predictor tool. The cache itself is a large collection of Perl storable files, each containing a large list of transcript entries.

Installation

Firstly, the script requires a modern version of Perl 5 to run. However you probably have such a version if you are running VEP, so this is likely not an issue

To install the script's dependencies, make sure cpanm is installed with

cpan App::cpanminus

and then install the dependencies using

cpanm --installdeps .

If you are using a system version of Perl, running these commands may give you an error message along these lines:

Can't write to /usr/local/share/perl/5.22.2 and /usr/local/bin

If this is the case, make sure to run these commands with sudo

You'll also need a copy of the VEP cache itself. You may already have one if you're already using VEP. In this case, simply use the directory of the cache as the first argument to the extract script. If not, you can use the downloader script to obtain it, and use vep_cache (a relative path) as the cache path (documented below).

Downloader Script

This script downloads a certain version of the VEP cache into the repository directory (the same directory as the script). Use it as follows:

Usage: download.sh -c cache_type -e ensembl_release -g genome-build
  -c, --cache-type
    The version of the cache to download. Either 'merged', 'ensembl' or 'refseq'
  -e, --ensembl-release
    The ensembl release number to download the cache for. e.g. 75, 85 etc.
  -g, --genome-build
    The grch build version to download for. e.g. 37, 38

Usage

Use the extract script as follows:

./extract.pl /path/to/cache path_1:column_1 path_2:column_2

In other words, the script's first argument is a directory path indicating where the cache is located.

The second and all following arguments are descriptors indicating which field you want to extract from the cache and what to name them. Each argument is a pair of path:column_name pairs. For example, this second argument might be _trans_exon_array.0.stable_id:exon_id (explained below)

• path: A dot separated string, where each segment is a hash key, indicating which field to choose. For example, "_trans_exon_array.0.stable_id" would choose the stable_id of the first exon in the transcript, which in json terms means extract transcript["_trans_exon_array"][0][stable_id] for each transcript. You might find it helpful to look at the sample.json file in the repository, as this will give you some indication as to which fields are stored in the VEP cache.
• column_name: A string indicating the name of the column to store this data in. For example, if we used the path above, along with the column_name of "exon_id", we would get output as follows:

exon_id
"id67641"
"id67642"
"NM_152663.3.1"
"XM_005245297.1.1"
"id67643"
"id67663"
"ENSESTE00000219503"
"ENSESTE00000220088"