Semantic technologies allow scientists to rapidly integrate data from experimental datasets and online databases. On the computational side, the massive scale of these datasets poses some challenges. We find that there are additional challenges on the human side. Problems we encounter in practice include translating human questions to SPARQL queries, adjusting data models and ontologies to match the problem space of scientists, visualising data, and providing user-friendly access.
We have integrated several Drosophila melanogaster datasets and built a public showcase. This showcase integrates diverse data types such as genes, proteins, FlyAtlas expression data and BioPAX pathways. This data is stored in a triple store running in a cloud instance. We provide several options for visualizing the data. First, a custom Cytoscape plugin can extract triple data and visualize it in network form. Secondly, a PathVisio plugin can be used to create visualisations on top of pathway diagrams derived from WikiPathways (See Fig. 1).
By following the demo, you should get a clear idea how linked data can be used to integrate diverse Drosophila datasets, how this can be visualized in pathway and network context, and how this can help to answer research questions.
The integrated data, SPARQL tutorial and videos, RDF downloads, cytoscape session and live SPARQL endpoint are all available at http://fly.cloud.generalbioinformatics.com/