Fig. 3From: A multi-task graph-clustering approach for chromosome conformation capture data sets identifies conserved modules of chromosomal interactionsOverview of Arboretum-Hi-C. Given n graphs of genomic regions, each representing a Hi-C data set in n species or cell lines, our method clusters the regions based on their interactions, while exploiting the relatedness (shown as the hierarchy) among the data sets. The figure shows an example of three Hi-C data sets with two clusters. The example shows that some regions keep their cluster assignments in all species, while other regions (diverged regions) change their cluster based on changes in their interactions with other regions. Dashed lines represent a mapping of regions from one data set to another. For cell lines of the same species, this mapping is trivial as the same regions are studied. For multiple species, this mapping requires one to find orthologous regions between speciesBack to article page