Hierarchical clustering may be performed on either the entire gene set or on a selected subset by selecting Clustering > Hierarchical.
The Hierarchical Clustering method groups data points by agglomerating them one-by-one into ever-growing groups. This grouping is done by first finding the shortest link among all of the data points, and then combining those two points into a group. The algorithm then finds the next shortest distance, including the distance from other points to this new group, and groups even further. Once all of the data points have been grouped, the resulting clusters are displayed in the Heat Map view with the associated Experiment Tree and Gene Tree on the horizontal and vertical axes, respectively.
Although best displayed in the Heat Map view, Hierarchical Clustering can also be displayed in the Line Thumbnails view by adjusting your Clustering Parameters.
By default, the Euclidean distance metric and the Centroid linkage method are used in Hierarchical Clustering. However, several distance metrics and linkage methods are available within ArrayStar and can be easily selected in the Clustering Parameters window.