Data Provenance Management of Bioinformatics Workflows in Federated Clouds

In Bioinformatics, the reproducibility of experiments is a fundamental principle to which the data provenance significantly contributes through the acquisition and management of information about the trajectory of the data in a workflow. In addition to the data provenance, aspects such as program configurations and the entire computational environment must be considered to achieve this goal. Cloud computing can provide computational resources, hiding technical details and providing an accessible and configurable on-demand environment for researchers. Cloud federation enables the broad distribution of services and a flexible combination of computing power. Considering this particular scenario, we propose a management platform that collects the provenance data of Bioinformatics workflows allowing portability between different clouds in federated clouds using the Infrastructure as a Service model. This platform is composed of tools that support running these workflows with data provenance captured in NoSQL databases. The retrospective data provenance is captured according to the PROV-DM norms. The findings indicate that the proposed platform can be used for some of the most commoninly available clouds.