Managing data provenance in genome project workflows

Abstract

In this article, we propose the application of the PROV-DM model to manage data provenance for workflows designed to support genome projects. This provenance model aims at storing details of each execution of the workflow, which include raw and produced data, computational tools and versions, parameters, and so on. This way, biologists can review details of a particular workflow execution, compare information generated among different executions, and plan new ones more efficiently. In addition, we have created a provenance simulator to facilitate the inclusion of a provenance data model in genome projects. In order to validate our proposal, we discuss a case study of an RNA-Seq project that aims to identify, measure and compare RNA expression levels across liver and kidney RNA samples produced by high-throughput automatic sequencers.