clustTV (MatArray Toolbox)

MatArray Toolbox

clustTV

clustTV exports a clustering created with hierarc so that it can be viewed with TreeView, the program developped by Michael Eisen.

Syntax

clustTV(filename,data,name,w,tG,tA)
clustTV(filename,data,name,w,tG,tA,geneDescr)

Description

The clusterings tG and tA, created with hierarc on the data data are exported to a TreeView compatible clustering with the name filename. Missing values in data should be set to NaN. In the case one or both clustering are absent, use [] as placeholders for tG and/or tA.

name is a cell array of strings used to indicate the names of the arrays.

The parameter w can be used to indicate which genes (i.e. lines) in the data matrix were used to determine the clusterings, and should be kept for the visualization. Use [] as a placeholder if it is not needed.

In the case the geneDescr parameter is present, it is used to link to a file giving the gene descriptions. Each line of the file should give the description of one gene. If geneDescr is a numerical parameter, a look-up table is used to determine the name of the gene description file (which should be somewhere in the path). If it is a char array, it is used directly as the name of the description file. In the case no parameter are given, the genes are designed by their line number in the matrix data. The numerical codes are chosen by default to correspond to NEN slides, that is

2400 genes arrays - '2400.txt'
4800 genes arrays - '4800.txt'
Oncogenes arrays - 'fa-3.txt'

This should be modified to suit other slides directly in the code of clustTV.

Examples

Creation of a random ratio matrix with 100 genes and 10 slides, and of a name char array with the slide names.

ratio = randn(100,10);
name = { 'a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'i' 'k' };

Say we only want to keep the genes where at least one absolute ratio is over 1.5

w = max(abs(ratio),[],2)>1.5;

Creation of an average linkage clustering on the genes where w == 1

tG = hierarc(1-corrcoef(ratio(w,:)'),1);

Export of the clustering with the name abc for visualization with TreeView ([] is used for tA, as no slide clustering is used)

clustTV('abc',ratio,name,w,tG,[]);

The clustering can be viewed by opening abc.cdt in the current directory with TreeView. The result will of course show little order as the ratio matrix was random.