Skip to main content
Fig. 6 | BMC Ecology and Evolution

Fig. 6

From: Investigating the reliability of molecular estimates of evolutionary time when substitution rates and speciation rates vary

Fig. 6

Flowchart of study design. Trees with branch lengths in units of time are simulated under three simulation conditions. For the Unlinked model, birth rates (\(\lambda\))and substitution rates (sub) vary independently, while for the Continuous model they are correlated. For the Punctuated model, trees are simulated as for Unlinked, then some of the branch length is redistributed to form bursts of substitutions occurring upon speciation. Trees and branch lengths are then used to simulate molecular sequence alignments. Time trees are then reconstructed using three different methods with two rate priors: theuncorrelated lognormal in BEAST 2 (UCLN), the autocorrelated lognormal in PAML (ACLN), and the uncorrelated lognormal in PAML (UCLN). Reconstructed trees are then compared to the original, simulated trees to determine the error (Median absolute % error, MAPE), uncertainty (HPD width) and bias towards over- or under-estimation (Gamma)

Back to article page