An Identity Crisis in the Life Sciences

Grid is an e-Science project assisting life scientists to build workflows that gather and co-ordinate data from distributed, autonomous, replicated and heterogeneous resources. The provenance logs of workflow executions are recorded as RDF graphs. The log of one workflow run is used to trace the history of its execution process; however, by aggregating provenance logs of workflow reruns, or runs of different workflows, we can gather the provenance of a common data product shared in multiple derivation paths. This aggregation relies on accurate and universal identification of each data product. The nature of bioinformatics data and services, however, makes this difficult. We describe the identity problem in bioinformatics data, and present a protocol for managing identity coreferences and allocating identity to collected and computed data products. The ability to overcome this problem means that the provenance of workflows in bioinformatics and other domains can be themselves exploited to enhance the practice of e-Science.