Online Ancient Genome Repository

[Christopher Franks]

Hi all

The University of Adelaide Libraries and Australian Centre for Ancient DNA this week launched the Online Ancient Genome Repository (OAGR), a installation of MyTardis making openly available a collection of ancient human DNA data: https://www.oagr.org.au/

Some key points about our use of MyTardis:

• We are using a customised MyTardis forked from the develop branch around December 2014 (before the 3.6 release)
• Data files are hosted on a Hitachi Content Platform server – essentially a big NAS/RAID machine – and accessed via a MyTardis storage box using django-storages and the S3 API, which is supported by the HCP for object store put/get/list/delete operations
• New Django models/database tables were added to store relational data about the provenance and production of ancient human DNA datasets published in OAGR. These are hooked into MyTardis via a relationship with the tardis_portal.Dataset model, a customised version of the MyTardis dataset view, and some additional frontend views (the Source index and detail views)
• The current deployment is running on Nginx + gunicorn, using PostgreSQL as the database backend and Elasticsearch for search functionality

OAGR is deployed to the NeCTAR cloud using the immutable server model. In addition to our data files being stored on the HCP server, the OAGR database/system state is also backed up to the HCP and our MyTardis fork, NeCTAR build process and configuration settings are hosted on GitHub. This separation of the system architecture into volatile/impermanent and persistent layers means it is straightforward to build new instances of OAGR, either for production or development/testing – for example, to redeploy the production instance to a larger size NeCTAR VM.

Head over to https://github.com/modc08 for goodies including our MyTardis fork, NeCTAR build/deployment scripts, and some Python scripts for ingesting data into OAGR through interaction with both the HCP object store (via S3 API/boto) and MyTardis (via the MyTardis REST API).

The obligatory acknowledgements: OAGR was developed by the Australian Centre for Ancient DNA, University of Adelaide Libraries and eResearch SA, as part of a project supported by the Australian National Data Service (ANDS). ANDS is supported by the Australian Government through the National Collaborative Research Infrastructure Strategy Program.

Many thanks also to Steve A and Grischa who provided some valuable advice during the course of the project.

Cheers

Christopher Franks
Open Data Project Officer
University Libraries
The University of Adelaide
SA 5005 Australia
8313 1062
christoph...@adelaide.edu.au

CRICOS Provider Number 00123M