GEMINI - integrative exploration of genetic variation and genome annotations.

Overview

The intent of GEMINI (GEnome MINIing) is to provide a simple, flexible, and powerful framework for exploring genetic variation for personal and medical genetics. GEMINI is unique in that it integrated genetic variation (from VCF files) with a wealth of genome annotations into a unified database framework. Using this integrated database as the analysis framework, we aim to leverage the expressive power of SQL for data analysis, while attempting to overcome the fundamental challenges associated with using databases for very large (e.g. 1,000,000 variants times 1,000 samples yields one billion genotypes) datasets.

Documentation

The official documentation is here: http://gemini.readthedocs.org/en/latest/

Mailing List: https://groups.google.com/forum/?fromgroups#!forum/gemini-variation

Acknowledgements

GEMINI is being developed in the Quinlan lab (quinlanlab.org) at the University of Virginia and is led by Uma Paila and Aaron Quinlan. Substantial contributions to the code base have also been made by Brad Chapman (@chapmanb) and Rory Kirchner (@roryk) at the Harvard School of Public Health.

Installation

Install GEMINI using the automated installation script, gemini_install.py. This script installs GEMINI along with required python libraries, third party tools and data files used for variant annotation. The installation documentation contains additional details on installed files and tools.::

wget https://raw.github.com/arq5x/gemini/master/gemini/scripts/gemini_install.py
python gemini_install.py /usr/local /usr/local/share/gemini

If you don't have administrative (or sudo) priveleges on you machine, use the --nosudo option::

python gemini_install.py --nosudo ~/src ~/data/gemini

If using the --nosudo option, you will need to add the gemini/bin directory that was created as part of the installation to your PATH.

The first directory specified (/usr/local in the first example above) dictates the directory in which the gemini software should be installed. The second directory specified (/usr/local/share/gemini in the first example above) specifies the directory in which the gemini annotations files should be installed. NOTE that these are rather large files (some > 1Gb).

License

GEMINI is freely available under the MIT license.