Fig. 16

A comparison of MLS and divergence from linearity for the number of clusters \(k=15\). We generated 1000 random (black), linear (green), and k-medoids clusterings using non-normalized Hi-C (blue) and k-medoids clusterings using normalized Hi-C (red). minimum linear (olive), optimal minimizing divergence from linearity (orange), and optimal maximizing divergence from linearity (brown) are also plotted. Mean divergence from linearity for blue k-medoids points is 5.3 and 16.5 for red k-medoids points