Workflows are useful ways to support scientific researchers in carrying out repetitive analytical tasks on digital information. Web services can provide a useful implementation mechanism for workflows, particularly when they are distributed, i.e., where some of the data or processing resources are remote from the scientist initiating the workflow. While many scientific workflows primarily involve operations on structured or numerical data, all interpretation of results is done in the context of related work in the field, as reflected in the scientific literature. Text mining technology can assist in automatically building helpful pathways into the relevant literature as part of a workflow in order to support the scientific discovery process. In this paper we demonstrate how these three technologies - workflows, text mining, and web services - can be fruitfully combined in order to support bioinformatics researchers investigating the genetic basis of two physiological disorders - Graves' disease and Williams syndrome.