3.3.4 Clustering profiles

Clustering is a general technique for unsupervised data classification widely used in microarray data analysis. A VAMP function offers the possibility to perform a hierarchical clustering (Kaufman and Rousseuw, 1990) on the array CGH profiles (see Figure 3.43).

Figure 3.43: Tools $\rightarrow$ Clustering $\rightarrow$ Compute - The user can open a new window of dialog for clustering.

The clustering can be performed on different variables (see Figure 3.44):

Probe LogRatio:

The Probe LogRatio values of the whole genomic profile are used

Probe Smoothing:

The Probe smoothing values (i.e. the results of a segmentation algorithm) of the whole genomic profile are used

Probe Status:

The Probe statuses (i.e. the results of a segmentation algorithm) of the whole genomic profile are used

Regions Status:

Regions either selected manually or identified by our algorithm (see section 3.3.3) are used

Exclude sexual chromosomes

.

Different options are available:

Distance metric:

Euclidian, Pearson and Manhattan distance are available

Group metric:

Ward, Single linkage, Group Average and Complete linkage are available

VAMP displays the results as a cluster view including a heat map and the trees resulting from the clustering algorithm (Figure 3.45).

Figure 3.44: Clustering profies - Different clustering options are available.

Figure 3.45: VAMP interface - Dotplot view of array-CGH profiles (middle panel), and dendrogram resulting from a hierarchical clustering (right panel). In between, color-coded clinical information about the samples, with a legend (bottom left). Data from Nakao et al. (2004)