Figure 4. For visual exploration and analysis of GTC data, our implementation provides a visual interface for selecting subsets of particles that meet a set of user-defined criteria. Here, “interesting” is defined as those data points that satisfy a set of multivariate range combinations via the parallel coordinates interface (lower image). The subset satisfying these range conditions then appears in the physical view (top image), where the view may be manipulated, the color transfer changed to draw attention to specific particles based upon other user-defined criteria, or subject to other types of visual or traditional analysis. (Image courtesy S. Ahern, Oak Ridge National Laboratory)

Figure 5. Here, we define three groups of “interesting” data – colored red, green, and blue – with a parallel coordinates interface (bottom pane), and those data points satisfying the multivariate range condition appear in the physical view (upper right). Here, we introduce a third linked view (upper left) that shows a 3D scatterplot – each of the data points from the three “groups of interesting” are colored red, green or blue (according to which “group of interesting” they belong). The three axes of the scatterplot are the expression levels for three specific genes; there are on the order of about 20 different gene expression levels per cell in this dataset – we picked three genes from the group of 20 for the purposes of display. This type of linked view presentation is very helpful in conveying different types of relationships in complex data. (Image courtesy of Oliver Rübel, Lawrence Berkeley National Laboratory)