Figure 1

Classification of biclusters of genes and probe sets (m27-1p). In each bicluster that exemplifies a particular class, a small circle located at the intersection of the horizontal and vertical lines indicates that the corresponding genes (vertical lines) are targeted by the corresponding probe sets (horizontal lines). In each panel, the probe set name without asterisk is the one used to find genes and probe sets which are related to it. The number of asterisks indicates the depth at which the recursive algorithm found the related gene(s) and probe set(s). A continuous line joins each probe set to the gene it is assigned to by PSAWNml (a thicker horizontal line indicates a probe set which targets a single gene). Interrupted lines join probe sets to all other targeted, but not assigned genes. Empty black and filled coloured round symbols at the intersection between horizontal (probe set) and vertical (gene) lines indicate, respectively, that the gene is targeted by the maximum possible number of probes, or by a smaller number of probes. For genes that are not targeted by the maximum possible number of probes, the change in line colour (from warm towards cool) indicates that the number of targeting probes decreases. If several probe sets are assigned to the same gene, as in class MS, MM, CX and HX, probe sets which are considered as similar by the PSAWN method are boxed by a large square of identical colour. Gene names starting with ‘Ens’ correspond to shortened Ensembl IDs (for instance, Ens78653 corresponds to ENSMUG00000078653); other names are AceView Ids.