Federated Galaxy: Biomedical Computing at the Frontier

Abstract

Biomedical data exploration requires integrative analyses of large datasets using a diverse ecosystem of tools. For more than a decade, the Galaxy project (https://galaxyproject.org) has provided researchers with a web-based, user-friendly, scalable data analysis framework complemented by a rich ecosystem of tools (https://usegalaxy.org/toolshed) used to perform genomic, proteomic, metabolomic, and imaging experiments. Galaxy can be deployed on the cloud (https://launch.usegalaxy.org), institutional computing clusters, and personal computers, or readily used on a number of public servers (e.g., https://usegalaxy.org). In this paper, we present our plan and progress towards creating Galaxy-as-a-Service-a federation of distributed data and computing resources into a panoptic analysis platform. Users can leverage a pool of public and institutional resources, in addition to plugging-in their private resources, helping answer the challenge of resource divergence across various Galaxy instances and enabling seamless analysis of biomedical data.

Keywords: cloud bursting; data federation; service computing.

Fig. 1.

The planned evolution of Galaxy from a set of fragmented instances to a service-based solution with dynamically pluggable infrastructure.

Fig. 2.

A tentative interface of linking additional resources by a user.

Fig. 3.

(A) Conceptual architecture of components making up GaaS, their matching technologies, and (B) simplified invocation flow.