**Cluster** is drawn as a rounded rectangle.
It encompasses more than one nodes inside.
Numbers at the top of the clusters indicate *the number of nodes* inside the cluster
Labels following these numbers are the property terms with the highest scores (score = the proportion of the term in the cluster x term's weight), which provide insight into the biological meaning of each cluster.
Based on the maximal and minimal numbers of members of all clusters in each view, the displayed sizes of the clusters are calculated as BASE+((N-MIN)/(MAX-MIN))*C where BASE = the base size of any clusters, N = the number of members of the cluster, MIN = the number of members of the smallest cluster among all clusters, MAX = the number of members of the largest cluster among all clusters, and C = an arbitrary constant.
**NOTE:** You can also force NaviCluster to display clusters with sizes proportional to the number of members of them via the parameter *-pp* in the command line:
*java –Xms1024m –jar NaviCluster.jar –pp*
The color saturations of the clusters also reflect the numbers of nodes inside.
Solid lines denote meta-edges, which indicate that the two clusters have edges between their members. |