Installation

This installation guide assumes the use of the BASH Unix shell and a 64-bit Linux system. gNALI may either be installed directly or as a Bioconda package. gNALI may either be run on a single machine or a computing cluster.

Bioconda

The simpliest and recommended way to install gNALI is using the Bioconda channel for the conda package management system. We recommend installing conda with the Miniconda package. Note that the choice of Miniconda only affects the Python version in root environment. We recommend installing Miniconda using Python 3.8 64-bit Linux installer.

Overview

The Bioconda-based gNALI installation involves the following:

  1. Installing Bioconda
  2. Installing the "gNALI" Bioconda package (conda install gnali).

We provide detailed instructions below.

Miniconda (Python 3.8)

Bioconda requires conda to be installed and we recommend using the Miniconda package. Miniconda may be installed with the follow instructions:

wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
chmod 755 Miniconda3-latest-Linux-x86_64.sh
./Miniconda3-latest-Linux-x86_64.sh

You will likely want Miniconda to append the install location to your PATH and will need to select this option during the installation process. After installation, you will then need to either open a new terminal or source your bashrc file in the current terminal for Miniconda to become available on the PATH:

source ~/.bashrc

You can check if your Miniconda installation was successful with the following:

conda --version

Bioconda

You will need to add the following channels to conda. They must be added in this order so that priority is set correctly.

conda config --add channels conda-forge
conda config --add channels defaults
conda config --add channels anaconda
conda config --add channels bioconda

gNALI (Miniconda 3.8)

The follwing instructions assume you are using the Python 3.8 version of Miniconda. In this circumstance, we need to install gNALI within a Python 3.7 environment:

conda create --name gnali-env python=3.7 gNALI

This gNALI environment can be activated with the following:

conda activate gnali-env

You can check if gNALI was installed correctly with the following:

gnali --version

After installing, optionally run the command gnali_get_data <reference genome> to download reference files required to add loss-of-function annotations. * For use with gnomADv2.1.1 or gnomADv3.1.1, you do not have to run gnali_get_data * For use with custom databases WITH loss-of-function annotations, you do not have to run gnali_get_data * For use with custom databases WITHOUT loss-of-function annotations, run gnali_get_data grch37 or gnali_get_data grch38 depending on the reference genome used

The files downloaded by gnali_get_data for each reference genome require abut 35GB of disk space and took 1.5 hours on our systems with 16GB of RAM and a 3.20GHz processor.

The current environment may be deactivated with the following:

conda deactivate

It is important to note that this gNALI Bioconda environment will need to be activated in order to run the gNALI application. However, the benefit is that your system will be shielded from the Python 3.7 installation required by gNALI.

GitHub

The following instructions describe how to install gNALI using GitHub.

Overview

The GitHub gNALI installation involves the following:

  1. Installing Python 3.6 or later
  2. Installing dependencies
  3. Installing gNALI

We provide more detailed instructions below.

Python

gNALI requires Python 3.6 or later. The following may check your Python version:

python --version

Dependencies

Install the following dependencies according to their instructions:

And the following dependencies using sudo apt install <package> (requires administrative privilages):

gNALI

To install gNALI, get the tarball for the latest release and use:

pip install <link-to-latest-release.tar.gz>

Or, download the latest release and use:

pip install /path/to/gnali

After installing, optionally run the command gnali_get_data <reference genome> to download reference files required to add loss-of-function annotations. * For use with gnomADv2.1.1 or gnomADv3.1.1, you do not have to run gnali_get_data * For use with custom databases WITH loss-of-function annotations, you do not have to run gnali_get_data * For use with custom databases WITHOUT loss-of-function annotations, run gnali_get_data grch37 or gnali_get_data grch38 depending on the reference genome used

The files downloaded by gnali_get_data for each reference genome require abut 35GB of disk space and took 1.5 hours on our systems with 16GB of RAM and a 3.20GHz processor.