Clustering data using matlab

Question

I'm trying to cluster my data. This is the example of my data:

genes   param1  param2  ...
gene1   0.224   -0.113  ...
gene2   -0.149  -0.934  ...

I have a thousand of genes and a hundred of parameters. I wanted to cluster my data by both genes and parameters and used clustergram for it. As there are a lot of genes it's very difficult to understand anything using a picture. Now I want to have a text-information of the 15-20 biggest clusters of genes in my data. I mean 15-20 lists of genes, that belong to different clusters. How can I do this? Thanks

This is the example of clustergram I have from my data:

There are vertical and horizontal dendrograms here. As there is a lot of rows, it's impossible to see anything on vertical dendrogram (I need only this one). As far as I understand, dendrogram creates a binary clusters from my data, and there are N-1 clusters from N rows of data.As these are binary clusters, there is one cluster, on the next step it splits into two, then again into two and so on. Can I get information about which genes are in which clusters on the 4-th step, for example, when there are 16 clusters?

Answer 1

To see interesting parts of the dendrogram and heatmap more clearly, you can use the zoom button on the toolbar to select regions of interest and zoom in on them.

To find out which genes/variables are in a particular cluster, right-click on a point in one of the dendrograms that represents the cluster you're interested in, and select Export to Workspace. You'll get a structure with the following fields:

1. GroupNames — Cell array of text strings containing the names of the row or column groups.
2. RowNodeNames — Cell array of text strings containing the names of the row nodes.
3. ColumnNodeNames — Cell array of text strings containing the names of the column nodes.
4. ExprValues — An M-by-N matrix of intensity values, where M and N are the number of row nodes and of column nodes respectively.