Skip to main content
Figure 11 | BMC Bioinformatics

Figure 11

From: A visual analytics approach for understanding biclustering results from microarray data

Figure 11

Bicluster visualization. Three of the biggest biclusters that Bimax algorithm found for microarray data discussed in the Results section. a) Biclusters represented by their wrapping hulls. This visualization is useful to draw an overview of how biclusters are related. b) Nodes are now represented as glyphs. Attention goes into intersecting areas, which can be seen as 'supergroups' of genes and/or conditions (x, y, z). These structures will frequently appear in more complex visualizations. c) Detail of b). The nodes in a supergroup are drawn as pie charts. These nodes have a strong relationship under the specific criteria of the biclustering algorithm. For example, in the case of (z), it means that eight genes have very high transcription levels with similar variations under two subsets of conditions (those from biclusters 2 and 3). These two subsets share one condition (square shaped), present in the supergroup. In addition, we can see that biclusters 1 and 2 share two conditions; also there are no conditions shared by the three biclusters; finally bicluster 2 is secondary, because it only has six genes and one condition not grouped in biclusters 1 or 3.

Back to article page