Skip to main content
. 2018 Sep 18;10(11):2882–2898. doi: 10.1093/gbe/evy200

Fig. 3.

Fig. 3.

—Distributions of the mean Ks of gene pairs comprising the components closest to the true age of the WGD at Ks = 0.5 and Ks = 1.0, when the number of components for each mixture model is optimized by ML. Horizontal black lines represent the true age of a WGD. Panels a–f represent WGDs at Ks = 0.5, and panels g–l are WGD at Ks = 1.0. Results for known simulated node-averaged Ks when a WGD is at Ks = 0.5 are shown in panels (a), (b), and (c) for the exponential+normal, exponential+log normal, and normal mixture model, respectively. Results for the WGD at Ks = 0.5 for estimated Ks are in panels (d), (e), and (f). The distributions of means of components closest to the true WGD age when the WGD is at Ks = 1.0 is given in panels (g), (h), and (i) for known simulated Ks, and panels (j), (k), and (l) for estimated Ks.