Navigation and Exploration of Interconnected Pathways

Proceedings of EuroVis 2008

Visualizing pathways, i. e. models of cellular functional networks, is a challenging task in computer assisted
biomedicine. Pathways are represented as large collections of interwoven graphs, with complex structures present
in both the individual graphs and their interconnections. This situation requires the development of novel visu-
alization techniques to allow efficient visual exploration. We present the Caleydo framework, which incorporates
a number of approaches to handle such pathways. Navigation in the network of pathways is facilitated by a hi-
erarchical approach which dynamically selects a working set of individual pathways for closer inspection. These
pathways are interactively rendered together with visual interconnections in a 2.5D view using graphics hard-
ware acceleration. The layout of individual graphs is not computed automatically, but taken from the KEGG
and BioCarta databases, which use layouts that life scientists are familiar with. Therefore they encode essential
meta-information. While the KEGG and BioCarta pathways use a pre-defined layout, interactions such as link-
ing+brushing, neighborhood search or detail on demand are still fully interactive in Caleydo. We have evaluated
Caleydo with pathologists working on the determination of unknown gene functions. Informal experiences confirm
that Caleydo is useful in both generating and validating such hypotheses.
Even though the presented techniques are applied to medical pathways, the proposed way of interaction is not
limited to cellular processes and therefore has the potential to open new possibilities in other fields of application.

PDF