Bicluster analysis. (A) A bicluster corresponds to a submatrix defined by row and column subsets. Both subsets are not known in advance. After reordering the original data matrix, it can be seen as the rectangle with the yellow border. (B) EXPANDER summarizes bicluster analysis results in a table that lists the dimensions (numbers of genes and conditions) of the biclusters identified and their scores. Clicking on a row in this table pops-up a window with the submatrix view of the selected bicluster. Below the table there are two examples of biclusters identified in a dataset comprising some 1,000 genes measured across over 70 conditions in human cells. Row and column labels are gene and condition names for the bicluster, respectively.
Shamir et al. BMC Bioinformatics 2005 6:232 doi:10.1186/1471-2105-6-232