A bioinformatics analysis pipeline is often highly elaborate, due to the inherent complexity of biological systems and the variety and size of datasets. A digital equivalent of the 'Materials and Methods' section in wet laboratory publications would be highly beneficial to bioinformatics, for evaluating evidence and examining data across related experiments, while introducing the potential to find associated resources and integrate them as data and services. We present initial steps towards preserving bioinformatics 'materials and methods' by exploiting the workflow paradigm for capturing the design of a data analysis pipeline, and RDF to link the workflow, its component services, run-time provenance, and a personalized biological interpretation of the results. An example shows the reproduction of the unique graph of an analysis procedure, its results, provenance, and personal interpretation of a text mining experiment. It links data from Taverna, myExperiment.org, BioCatalogue.org, and ConceptWiki.org. The approach is relatively 'light-weight' and unobtrusive to bioinformatics users. © 2010 Springer-Verlag.
Roos, M., Bechhofer, S., Zhao, J., Missier, P., Newman, D. R., De Roure, D., & Marshall, M. S. (2010). A linked data approach to sharing workflows and workflow results. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 6415 LNCS, pp. 340–354). https://doi.org/10.1007/978-3-642-16558-0_29