Gene expression clusters (cut into flat groups)
A hierarchical-clustering dendrogram of gene-expression samples, sliced at a chosen height so the continuous tree collapses into distinctly colored flat clusters with a dashed threshold line.
For the computational biologist
What this shows
The same tree the phylo engine draws for evolution, read instead as the output of hierarchical agglomerative clustering (Sokal & Michener 1958 — the form scipy.cluster.hierarchy.dendrogram produces). Each internal node sits at its merge height: the cophenetic distance at which its two child clusters fuse. Leaves align on a common baseline, branches are rectangular elbows, and a height axis lets you read off exactly how far apart any two leaves are before they first share a cluster — (A, B) join low and tight, while (D, E) only merge with the rest near the top.
The cut 4 line is what makes this actionable. It slices the tree at height 4: every subtree that fuses below that line becomes one flat cluster, each painted a distinct color, with a dashed threshold line drawn across the chart. This is the dendrogram counterpart of choosing k clusters in fcluster — turning a continuous similarity tree into the discrete groups you actually act on.