Figure 1

Enrichment of clusters with genes from the same chromosome. The dendrograms resulting from standard flexible beta and GDEC clustering were cut at 200 clusters. The chromosomal locations of the genes in each cluster were tabulated, and for each cluster the number of genes from the most abundantly represented chromosome was scored as a fraction of the number of genes in the cluster. These proportion scores are plotted in ranked order from left to right for each of the three largest datasets. The red lines indicate GDEC clustering, the green lines standard flexible beta clustering and the blue lines GDEC clustering after destruction of correlation structure by permutation. The single black line indicates the profile obtained by randomly assigning genes to clusters.