Skip to main content
. 2018 Mar 20;35(7):1770–1782. doi: 10.1093/molbev/msy044

Fig. 1.

Fig. 1.

(a) Computational time taken by RelTime and MCMCTree (Bayesian method) to estimate divergence times for data sets containing increasing number of sequences (n). The tested sequence alignment consisted of 4,493 sites in which sequence evolution was simulated with extensive rate variation (RR50 data from Tamura et al. 2012). RelTime’s speed advantage increases with data volume by O(n2). (b) Distribution of computation speed ratio of RelTime to MCMCTree for 70 data sets, each containing 100 ingroup sequences that were simulated as described in the Materials and Methods.