Fig. 2
From: Finding high posterior density phylogenies by systematically extending a directed acyclic graph
An example showing how the sDAG is more flexible than the conditional clade distribution of [2, 5, 8]. Specifically, we may have different splitting probabilities for the clade \(\{2,3,4\}\) depending on which subsplit it is contained in (either \(\{\{0\},\{2, 3, 4\}\}\) or \(\{\{1\},\{2, 3, 4\}\}\)), showing that our approach is a strict generalization of clade-conditional approaches