Skip to main content
. 2017 Oct 11;284(1864):20170986. doi: 10.1098/rspb.2017.0986

Figure 1.

Figure 1.

Topological accuracy of ML versus Bayesian reconstructions for the most discordant comparison identified by PEA (see the text). D measures how much larger ML distances are from the true tree (dML) than are Bayesian distances (dBMR). MLE trees are indeed less accurate than BMRs (DMLE; mean = 17.63), but when conventional bootstrap thresholds are employed the difference in efficacy disappears (DML50; mean = 0.43).