Hierarchical clustering of features

A hierarchical clustering of features is a tree presentation of the similarity in expression profiles of the features over a set of samples (or groups).

The tree structure is generated by

  1. letting each feature be a cluster
  2. calculating pairwise distances between all clusters
  3. joining the two closest clusters into one new cluster
  4. iterating 2-3 until there is only one cluster left (which will contain all samples).
The tree is drawn so that the distances between clusters are reflected by the lengths of the branches in the tree. Thus, features with expression profiles that closely resemble each other have short distances between them, those that are more different, are placed further apart.

To start the clustering of features:

        Toolbox | Transcriptomics Analysis (Image expressionfolder)| Feature Clustering | Hierarchical Clustering of Features (Image feature_clustering)

Select at least two samples ( (Image array) or (Image rnaseq)) or an experiment (Image experiment).

Note! If your data contains many features, the clustering will take very long time and could make your computer unresponsive. It is recommended to perform this analysis on a subset of the data (which also makes it easier to make sense of the clustering. Typically, you will want to filter away the features that are thought to represent only noise, e.g. those with mostly low values, or with little difference between the samples). See how to create a sub-experiment in Creating sub-experiment from selection.

Clicking Next will display a dialog as shown in figure 28.99. The hierarchical clustering algorithm requires that you specify a distance measure and a cluster linkage. The distance measure is used specify how distances between two features should be calculated. The cluster linkage specifies how you want the distance between two clusters, each consisting of a number of features, to be calculated.

Image feature_clustering_step2
Figure 28.99: Parameters for hierarchical clustering of features.

At the top, you can choose three kinds of Distance measures:

Next, you can select different ways to calculate distances between clusters. The possible cluster linkage to use are:

At the bottom, you can select which values to cluster (see Selecting transformed and normalized values for analysis). Click Next if you wish to adjust how to handle the results. If not, click Finish.



Subsections